20181120.md

ArXiv cs.CV --Tue, 20 Nov 2018

1.OrthoSeg: A Deep Multimodal Convolutional Neural Network for Semantic Segmentation of Orthoimagery pdf

This paper addresses the task of semantic segmentation of orthoimagery using multimodal data e.g. optical RGB, infrared and digital surface model. We propose a deep convolutional neural network architecture termed OrthoSeg for semantic segmentation using multimodal, orthorectified and coregistered data. We also propose a training procedure for supervised training of OrthoSeg. The training procedure complements the inherent architectural characteristics of OrthoSeg for preventing complex co-adaptations of learned features, which may arise due to probable high dimensionality and spatial correlation in multimodal and/or multispectral coregistered data. OrthoSeg consists of parallel encoding networks for independent encoding of multimodal feature maps and a decoder designed for efficiently fusing independently encoded multimodal feature maps. A softmax layer at the end of the network uses the features generated by the decoder for pixel-wise classification. The decoder fuses feature maps from the parallel encoders locally as well as contextually at multiple scales to generate per-pixel feature maps for final pixel-wise classification resulting in segmented output. We experimentally show the merits of OrthoSeg by demonstrating state-of-the-art accuracy on the ISPRS Potsdam 2D Semantic Segmentation dataset. Adaptability is one of the key motivations behind OrthoSeg so that it serves as a useful architectural option for a wide range of problems involving the task of semantic segmentation of coregistered multimodal and/or multispectral imagery. Hence, OrthoSeg is designed to enable independent scaling of parallel encoder networks and decoder network to better match application requirements, such as the number of input channels, the effective field-of-view, and model capacity.

2.Event-Based Features Selection and Tracking from Intertwined Estimation of Velocity and Generative Contours pdf

This paper presents a new event-based method for detecting and tracking features from the output of an event-based camera. Unlike many tracking algorithms from the computer vision community, this process does not aim for particular predefined shapes such as corners. It relies on a dual intertwined iterative continuous -- pure event-based -- estimation of the velocity vector and a bayesian description of the generative feature contours. By projecting along estimated speeds updated for each incoming event it is possible to identify and determine the spatial location and generative contour of the tracked feature while iteratively updating the estimation of the velocity vector. Results on several environments are shown taking into account large variations in terms of luminosity, speed, nature and size of the tracked features. The usage of speed instead of positions allows for a much faster feedback allowing for very fast convergence rates.

3.Novel approach to locate region of interest in mammograms for Breast cancer pdf

Locating region of interest for breast cancer masses in the mammographic image is a challenging problem in medical image processing. In this research work, the keen idea is to efficiently extract suspected mass region for further examination. In particular to this fact breast boundary segmentation on sliced rgb image using modified intensity based approach followed by quad tree based division to spot out suspicious area are proposed in the paper. To evaluate the performance DDSM standard dataset are experimented and achieved acceptable accuracy.

4.Deeper Interpretability of Deep Networks pdf

Deep Convolutional Neural Networks (CNNs) have been one of the most influential recent developments in computer vision, particularly for categorization. There is an increasing demand for explainable AI as these systems are deployed in the real world. However, understanding the information represented and processed in CNNs remains in most cases challenging. Within this paper, we explore the use of new information theoretic techniques developed in the field of neuroscience to enable novel understanding of how a CNN represents information. We trained a 10-layer ResNet architecture to identify 2,000 face identities from 26M images generated using a rigorously controlled 3D face rendering model that produced variations of intrinsic (i.e. face morphology, gender, age, expression and ethnicity) and extrinsic factors (i.e. 3D pose, illumination, scale and 2D translation). With our methodology, we demonstrate that unlike human's network overgeneralizes face identities even with extreme changes of face shape, but it is more sensitive to changes of texture. To understand the processing of information underlying these counterintuitive properties, we visualize the features of shape and texture that the network processes to identify faces. Then, we shed a light into the inner workings of the black box and reveal how hidden layers represent these features and whether the representations are invariant to pose. We hope that our methodology will provide an additional valuable tool for interpretability of CNNs.

5.Event-based Gesture Recognition with Dynamic Background Suppression using Smartphone Computational Capabilities pdf

This paper introduces a framework of gesture recognition operating on the output of an event based camera using the computational resources of a mobile phone. We will introduce a new development around the concept of time-surfaces modified and adapted to run on the limited computational resources of a mobile platform. We also introduce a new method to remove dynamically backgrounds that makes full use of the high temporal resolution of event-based cameras. We assess the performances of the framework by operating on several dynamic scenarios in uncontrolled lighting conditions indoors and outdoors. We also introduce a new publicly available event-based dataset for gesture recognition selected through a clinical process to allow human-machine interactions for the visually-impaired and the elderly. We finally report comparisons with prior works that tackled event-based gesture recognition reporting comparable if not superior results if taking into account the limited computational and memory constraints of the used hardware.

6.DeepIR: A Deep Semantics Driven Framework for Image Retargeting pdf

We present \emph{Deep Image Retargeting} (\emph{DeepIR}), a coarse-to-fine framework for content-aware image retargeting. Our framework first constructs the semantic structure of input image with a deep convolutional neural network. Then a uniform re-sampling that suits for semantic structure preserving is devised to resize feature maps to target aspect ratio at each feature layer. The final retargeting result is generated by coarse-to-fine nearest neighbor field search and step-by-step nearest neighbor field fusion. We empirically demonstrate the effectiveness of our model with both qualitative and quantitative results on widely used RetargetMe dataset.

7.Deep Shape-from-Template: Wide-Baseline, Dense and Fast Registration and Deformable Reconstruction from a Single Image pdf

We present Deep Shape-from-Template (DeepSfT), a novel Deep Neural Network (DNN) method for solving real-time automatic registration and 3D reconstruction of a deformable object viewed in a single monocular image.DeepSfT advances the state-of-the-art in various aspects. Compared to existing DNN SfT methods, it is the first fully convolutional real-time approach that handles an arbitrary object geometry, topology and surface representation. It also does not require ground truth registration with real data and scales well to very complex object models with large numbers of elements. Compared to previous non-DNN SfT methods, it does not involve numerical optimization at run-time, and is a dense, wide-baseline solution that does not demand, and does not suffer from, feature-based matching. It is able to process a single image with significant deformation and viewpoint changes, and handles well the core challenges of occlusions, weak texture and blur. DeepSfT is based on residual encoder-decoder structures and refining blocks. It is trained end-to-end with a novel combination of supervised learning from simulated renderings of the object model and semi-supervised automatic fine-tuning using real data captured with a standard RGB-D camera. The cameras used for fine-tuning and run-time can be different, making DeepSfT practical for real-world use. We show that DeepSfT significantly outperforms state-of-the-art wide-baseline approaches for non-trivial templates, with quantitative and qualitative evaluation.

8.Explicit Bias Discovery in Visual Question Answering Models pdf

Researchers have observed that Visual Question Answering (VQA) models tend to answer questions by learning statistical biases in the data. For example, their answer to the question "What is the color of the grass?" is usually "Green", whereas a question like "What is the title of the book?" cannot be answered by inferring statistical biases. It is of interest to the community to explicitly discover such biases, both for understanding the behavior of such models, and towards debugging them. Our work address this problem. In a database, we store the words of the question, answer and visual words corresponding to regions of interest in attention maps. By running simple rule mining algorithms on this database, we discover human-interpretable rules which give us unique insight into the behavior of such models. Our results also show examples of unusual behaviors learned by models in attempting VQA tasks.

9.Modeling Local Geometric Structure of 3D Point Clouds using Geo-CNN pdf

Recent advances in deep convolutional neural networks (CNNs) have motivated researchers to adapt CNNs to directly model points in 3D point clouds. Modeling local structure has been proven to be important for the success of convolutional architectures, and researchers exploited the modeling of local point sets in the feature extraction hierarchy. However, limited attention has been paid to explicitly model the geometric structure amongst points in a local region. To address this problem, we propose Geo-CNN, which applies a generic convolution-like operation dubbed as GeoConv to each point and its local neighborhood. Local geometric relationships among points are captured when extracting edge features between the center and its neighboring points. We first decompose the edge feature extraction process onto three orthogonal bases, and then aggregate the extracted features based on the angles between the edge vector and the bases. This encourages the network to preserve the geometric structure in Euclidean space throughout the feature extraction hierarchy. GeoConv is a generic and efficient operation that can be easily integrated into 3D point cloud analysis pipelines for multiple applications. We evaluate Geo-CNN on ModelNet40 and KITTI and achieve state-of-the-art performance.

10.A Multi-Task Learning & Generation Framework: Valence-Arousal, Action Units & Primary Expressions pdf

Over the past few years many research efforts have been devoted to the field of affect analysis. Various approaches have been proposed for: i) discrete emotion recognition in terms of the primary facial expressions; ii) emotion analysis in terms of facial Action Units (AUs), assuming a fixed expression intensity; iii) dimensional emotion analysis, in terms of valence and arousal (VA). These approaches can only be effective, if they are developed using large, appropriately annotated databases, showing behaviors of people in-the-wild, i.e., in uncontrolled environments. Aff-Wild has been the first, large-scale, in-the-wild database (including around 1,200,000 frames of 300 videos), annotated in terms of VA. In the vast majority of existing emotion databases, their annotation is limited to either primary expressions, or valence-arousal, or action units. In this paper, we first annotate a part (around $234,000$ frames) of the Aff-Wild database in terms of $8$ AUs and another part (around $288,000$ frames) in terms of the $7$ basic emotion categories, so that parts of this database are annotated in terms of VA, as well as AUs, or primary expressions. Then, we set up and tackle multi-task learning for emotion recognition, as well as for facial image generation. Multi-task learning is performed using: i) a deep neural network with shared hidden layers, which learns emotional attributes by exploiting their inter-dependencies; ii) a discriminator of a generative adversarial network (GAN). On the other hand, image generation is implemented through the generator of the GAN. For these two tasks, we carefully design loss functions that fit the examined set-up. Experiments are presented which illustrate the good performance of the proposed approach when applied to the new annotated parts of the Aff-Wild database.

11.Aff-Wild2: Extending the Aff-Wild Database for Affect Recognition pdf

Automatic understanding of human affect using visual signals is a problem that has attracted significant interest over the past 20 years. However, human emotional states are quite complex. To appraise such states displayed in real-world settings, we need expressive emotional descriptors that are capable of capturing and describing this complexity. The circumplex model of affect, which is described in terms of valence (i.e., how positive or negative is an emotion) and arousal (i.e., power of the activation of the emotion), can be used for this purpose. Recent progress in the emotion recognition domain has been achieved through the development of deep neural architectures and the availability of very large training databases. To this end, Aff-Wild has been the first large-scale "in-the-wild" database, containing around 1,200,000 frames. In this paper, we build upon this database, extending it with 260 more subjects and 1,413,000 new video frames. We call the union of Aff-Wild with the additional data, Aff-Wild2. The videos are downloaded from Youtube and have large variations in pose, age, illumination conditions, ethnicity and profession. Both database-specific as well as cross-database experiments are performed in this paper, by utilizing the Aff-Wild2, along with the RECOLA database. The developed deep neural architectures are based on the joint training of state-of-the-art convolutional and recurrent neural networks with attention mechanism; thus exploiting both the invariant properties of convolutional features, while modeling temporal dynamics that arise in human behaviour via the recurrent layers. The obtained results show premise for utilization of the extended Aff-Wild, as well as of the developed deep neural architectures for visual analysis of human behaviour in terms of continuous emotion dimensions.

12.Addressing the Invisible: Street Address Generation for Developing Countries with Deep Learning pdf

More than half of the world's roads lack adequate street addressing systems. Lack of addresses is even more visible in daily lives of people in developing countries. We would like to object to the assumption that having an address is a luxury, by proposing a generative address design that maps the world in accordance with streets. The addressing scheme is designed considering several traditional street addressing methodologies employed in the urban development scenarios around the world. Our algorithm applies deep learning to extract roads from satellite images, converts the road pixel confidences into a road network, partitions the road network to find neighborhoods, and labels the regions, roads, and address units using graph- and proximity-based algorithms. We present our results on a sample US city, and several developing cities, compare travel times of users using current ad hoc and new complete addresses, and contrast our addressing solution to current industrial and open geocoding alternatives.

13.Handwriting Recognition of Historical Documents with few labeled data pdf

Historical documents present many challenges for offline handwriting recognition systems, among them, the segmentation and labeling steps. Carefully annotated textlines are needed to train an HTR system. In some scenarios, transcripts are only available at the paragraph level with no text-line information. In this work, we demonstrate how to train an HTR system with few labeled data. Specifically, we train a deep convolutional recurrent neural network (CRNN) system on only 10% of manually labeled text-line data from a dataset and propose an incremental training procedure that covers the rest of the data. Performance is further increased by augmenting the training set with specially crafted multiscale data. We also propose a model-based normalization scheme which considers the variability in the writing scale at the recognition phase. We apply this approach to the publicly available READ dataset. Our system achieved the second best result during the ICDAR2017 competition.

14.Injecting and removing malignant features in mammography with CycleGAN: Investigation of an automated adversarial attack using neural networks pdf

$\textbf{Purpose}$ To train a cycle-consistent generative adversarial network (CycleGAN) on mammographic data to inject or remove features of malignancy, and to determine whether these AI-mediated attacks can be detected by radiologists. $\textbf{Material and Methods}$ From the two publicly available datasets, BCDR and INbreast, we selected images from cancer patients and healthy controls. An internal dataset served as test data, withheld during training. We ran two experiments training CycleGAN on low and higher resolution images ($256 \times 256$ px and $512 \times 408$ px). Three radiologists read the images and rated the likelihood of malignancy on a scale from 1-5 and the likelihood of the image being manipulated. The readout was evaluated by ROC analysis (Area under the ROC curve = AUC). $\textbf{Results}$ At the lower resolution, only one radiologist exhibited markedly lower detection of cancer (AUC=0.85 vs 0.63, p=0.06), while the other two were unaffected (0.67 vs. 0.69 and 0.75 vs. 0.77, p=0.55). Only one radiologist could discriminate between original and modified images slightly better than guessing/chance (0.66, p=0.008). At the higher resolution, all radiologists showed significantly lower detection rate of cancer in the modified images (0.77-0.84 vs. 0.59-0.69, p=0.008), however, they were now able to reliably detect modified images due to better visibility of artifacts (0.92, 0.92 and 0.97). $\textbf{Conclusion}$ A CycleGAN can implicitly learn malignant features and inject or remove them so that a substantial proportion of small mammographic images would consequently be misdiagnosed. At higher resolutions, however, the method is currently limited and has a clear trade-off between manipulation of images and introduction of artifacts.

15.Contextual Face Recognition with a Nested-Hierarchical Nonparametric Identity Model pdf

Current face recognition systems typically operate via classification into known identities obtained from supervised identity annotations. There are two problems with this paradigm: (1) current systems are unable to benefit from often abundant unlabelled data; and (2) they equate successful recognition with labelling a given input image. Humans, on the other hand, regularly perform identification of individuals completely unsupervised, recognising the identity of someone they have seen before even without being able to name that individual. How can we go beyond the current classification paradigm towards a more human understanding of identities? In previous work, we proposed an integrated Bayesian model that coherently reasons about the observed images, identities, partial knowledge about names, and the situational context of each observation. Here, we propose extensions of the contextual component of this model, enabling unsupervised discovery of an unbounded number of contexts for improved face recognition.

16.Past, Present, and Future Approaches Using Computer Vision for Animal Re-Identification from Camera Trap Data pdf

The ability of a researcher to re-identify (re-ID) an individual animal upon re-encounter is fundamental for addressing a broad range of questions in the study of ecosystem function, community and population dynamics, and behavioural ecology. In this review, we describe a brief history of camera traps for re-ID, present a collection of computer vision feature engineering methodologies previously used for animal re-ID, provide an introduction to the underlying mechanisms of deep learning relevant to animal re-ID, highlight the success of deep learning methods for human re-ID, describe the few ecological studies currently utilizing deep learning for camera trap analyses, and our predictions for near future methodologies based on the rapid development of deep learning methods. By utilizing novel deep learning methods for object detection and similarity comparisons, ecologists can extract animals from an image/video data and train deep learning classifiers to re-ID animal individuals beyond the capabilities of a human observer. This methodology will allow ecologists with camera/video trap data to re-identify individuals that exit and re-enter the camera frame. Our expectation is that this is just the beginning of a major trend that could stand to revolutionize the analysis of camera trap data and, ultimately, our approach to animal ecology.

17.M2U-Net: Effective and Efficient Retinal Vessel Segmentation for Resource-Constrained Environments pdf

In this paper, we present a novel neural network architecture for retinal vessel segmentation that improves over the state of the art on two benchmark datasets, is the first to run in real time on high resolution images, and its small memory and processing requirements make it deployable in mobile and embedded systems.
The M2U-Net has a new encoder-decoder architecture that is inspired by the U-Net. It adds pretrained components of MobileNetV2 in the encoder part and novel contractive bottleneck blocks in the decoder part that, combined with bilinear upsampling, drastically reduce the parameter count to 0.55M compared to 31.03M in the original U-Net.
We have evaluated its performance against a wide body of previously published results on three public datasets. On two of them, the M2U-Net achieves new state-of-the-art performance by a considerable margin. When implemented on a GPU, our method is the first to achieve real-time inference speeds on high-resolution fundus images. We also implemented our proposed network on an ARM-based embedded system where it segments images in between 0.6 and 15 sec, depending on the resolution. Thus, the M2U-Net enables a number of applications of retinal vessel structure extraction, such as early diagnosis of eye diseases, retinal biometric authentication systems, and robot assisted microsurgery.

18.Do Normalization Layers in a Deep ConvNet Really Need to Be Distinct? pdf

Yes, they do. This work investigates a perspective for deep learning: whether different normalization layers in a ConvNet require different normalizers. This is the first step towards understanding this phenomenon. We allow each convolutional layer to be stacked before a switchable normalization (SN) that learns to choose a normalizer from a pool of normalization methods. Through systematic experiments in ImageNet, COCO, Cityscapes, and ADE20K, we answer three questions: (a) Is it useful to allow each normalization layer to select its own normalizer? (b) What impacts the choices of normalizers? (c) Do different tasks and datasets prefer different normalizers? Our results suggest that (1) using distinct normalizers improves both learning and generalization of a ConvNet; (2) the choices of normalizers are more related to depth and batch size, but less relevant to parameter initialization, learning rate decay, and solver; (3) different tasks and datasets have different behaviors when learning to select normalizers.

19.FD-GAN: Face-demorphing generative adversarial network for restoring accomplice's facial image pdf

Face morphing attack is proved to be a serious threat to the existing face recognition systems. Although a few face morphing detection methods have been put forward, the face morphing accomplice's facial restoration remains a challenging problem. In this paper, a face-demorphing generative adversarial network (FD-GAN) is proposed to restore the accomplice's facial image. It utilizes a symmetric dual network architecture and two levels of restoration losses to separate the identity feature of the morphing accomplice. By exploiting the captured face image (containing the criminal's identity) from the face recognition system and the morphed image stored in the e-passport system (containing both criminal and accomplice's identities), the FD-GAN can effectively restore the accomplice's facial image. Experimental results and analysis demonstrate the effectiveness of the proposed scheme. It has great potential to be implemented for detecting the face morphing accomplice in a real identity verification scenario.

20.Intention Oriented Image Captions with Guiding Objects pdf

Although existing image caption models can produce promising results using recurrent neural networks (RNNs), it is difficult to guarantee that an object we care about is contained in generated descriptions, for example in the case that the object is inconspicuous in image. Problems become even harder when these objects did not appear in training stage. In this paper, we propose a novel approach for generating image captions with guiding objects (CGO). The CGO constrains the model to involve a human-concerned object, when the object is in the image, in the generated description while maintaining fluency. Instead of generating the sequence from left to right, we start description with a selected object and generate other parts of the sequence based on this object. To achieve this, we design a novel framework combining two LSTMs in opposite directions. We demonstrate the characteristics of our method on MSCOCO to generate descriptions for each detected object in images. With CGO, we can extend the ability of description to the objects being neglected in image caption labels and provide a set of more comprehensive and diverse descriptions for an image. CGO shows obvious advantages when applied to the task of describing novel objects. We show experiment results on both MSCOCO and ImageNet datasets. Evaluations show that our method outperforms the state-of-the-art models in the task with average F1 75.8, leading to better descriptions in terms of both content accuracy and fluency.

21.Watermark Retrieval from 3D Printed Objects via Convolutional Neural Networks pdf

We present a method for reading digital data embedded in planar 3D printed surfaces. The data are organised in binary arrays and embedded as surface textures in a way inspired by QR codes. At the core of the retrieval method lies a Convolutional Neural Network, outputting a confidence map of the location of the surface textures encoding value 1 bits. Subsequently, the bit array is retrieved through a series of simple image processing and statistical operations applied on the confidence map. Extensive experimentation with images captured from various camera views, under various illumination conditions and from objects printed with various material colours, shows that the proposed method generalizes well and achieves the level of accuracy required in practical applications.

22.Collaborative Dense SLAM pdf

In this paper, we present a new system for live collaborative dense surface reconstruction. Cooperative robotics, multi participant augmented reality and human-robot interaction are all examples of situations where collaborative mapping can be leveraged for greater agent autonomy. Our system builds on ElasticFusion to allow a number of cameras starting with unknown initial relative positions to maintain local maps utilising the original algorithm. Carrying out visual place recognition across these local maps the system can identify when two maps overlap in space, providing an inter-map constraint from which the system can derive the relative poses of the two maps. Using these resulting pose constraints, our system performs map merging, allowing multiple cameras to fuse their measurements into a single shared reconstruction. The advantage of this approach is that it avoids replication of structures subsequent to loop closures, where multiple cameras traverse the same regions of the environment. Furthermore, it allows cameras to directly exploit and update regions of the environment previously mapped by other cameras within the system. We provide both quantitative and qualitative analyses using the synthetic ICL-NUIM dataset and the real-world Freiburg dataset including the impact of multi-camera mapping on surface reconstruction accuracy, camera pose estimation accuracy and overall processing time. We also include qualitative results in the form of sample reconstructions of room sized environments with up to 3 cameras undergoing intersecting and loopy trajectories.

23.SEIGAN: Towards Compositional Image Generation by Simultaneously Learning to Segment, Enhance, and Inpaint pdf

We present a novel approach to image manipulation and understanding by simultaneously learning to segment object masks, paste objects to another background image, and remove them from original images. For this purpose, we develop a novel generative model for compositional image generation, SEIGAN (Segment-Enhance-Inpaint Generative Adversarial Network), which learns these three operations together in an adversarial architecture with additional cycle consistency losses. To train, SEIGAN needs only bounding box supervision and does not require pairing or ground truth masks. SEIGAN produces better generated images (evaluated by human assessors) than other approaches and produces high-quality segmentation masks, improving over other adversarially trained approaches and getting closer to the results of fully supervised training.

24.ATOM: Accurate Tracking by Overlap Maximization pdf

While recent years have witnessed astonishing improvements in visual tracking robustness, the advancements in tracking accuracy have been severely limited. As the focus has been directed towards the development of powerful classifiers, the problem of accurate target state estimation has been largely overlooked. Instead, the majority of methods resort to simple multi-scale search in order to estimate the target bounding box. We argue that this approach is fundamentally limited as target estimation is a complex task, requiring high-level knowledge about the object. We thus address the problem of target state estimation in tracking.
We propose a novel tracking architecture consisting of dedicated target estimation and classification components. Due to the complex nature of target estimation, we propose a component that can be entirely trained offline on large-scale datasets. Our target estimation component is trained to predict the overlap between the target object and an estimated bounding box. By carefully integrating target-specific information in the prediction, our approach achieves previously unseen bounding box accuracy. Furthermore, we integrate a classification component that is trained online to guarantee high discriminative power in the presence of distractors. Our final tracking framework, comprised of a unified multi-task architecture, sets a new state-of-the-art on four challenging benchmarks. On the large-scale TrackingNet dataset, our tracker ATOM achieves a relative gain of 15%, while running at over 30 FPS.

25.Beyond Attributes: Adversarial Erasing Embedding Network for Zero-shot Learning pdf

In this paper, an adversarial erasing embedding network with the guidance of high-order attributes (AEEN-HOA) is proposed for going further to solve the challenging ZSL/GZSL task. AEEN-HOA consists of two branches, i.e., the upper stream is capable of erasing some initially discovered regions, then the high-order attribute supervision is incorporated to characterize the relationship between the class attributes. Meanwhile, the bottom stream is trained by taking the current background regions to train the same attribute. As far as we know, it is the first time of introducing the erasing operations into the ZSL task. In addition, we first propose a class attribute activation map for the visualization of ZSL output, which shows the relationship between class attribute feature and attention map. Experiments on four standard benchmark datasets demonstrate the superiority of AEEN-HOA framework.

26.Weakly Supervised Soft-detection-based Aggregation Method for Image Retrieval pdf

In recent year, the compact representations based on activations of Convolutional Neural Network (CNN) achieve remarkable performance in image retrieval. Some interested object only takes up a small part of the whole image. Therefore, it is significant to extract the discriminative representations that contain regional information of pivotal small object. In this paper, we propose a novel weakly supervised soft-detection-based aggregation (SDA) method free from bounding box annotations for image retrieval. In order to highlight the certain discriminative pattern of objects and suppress the noise of background, we employ trainable soft region proposals that indicate the probability of interested object and reflect the significance of candidate regions.
We conduct comprehensive experiments on standard image retrieval datasets. Our weakly supervised SDA method achieves state-of-the-art performance on most benchmarks. The results demonstrate that the proposed SDA method is effective for image retrieval.

27.Self-Referenced Deep Learning pdf

Knowledge distillation is an effective approach to transferring knowledge from a teacher neural network to a student target network for satisfying the low-memory and fast running requirements in practice use. Whilst being able to create stronger target networks compared to the vanilla non-teacher based learning strategy, this scheme needs to train additionally a large teacher model with expensive computational cost. In this work, we present a Self-Referenced Deep Learning (SRDL) strategy. Unlike both vanilla optimisation and existing knowledge distillation, SRDL distils the knowledge discovered by the in-training target model back to itself to regularise the subsequent learning procedure therefore eliminating the need for training a large teacher model. SRDL improves the model generalisation performance compared to vanilla learning and conventional knowledge distillation approaches with negligible extra computational cost. Extensive evaluations show that a variety of deep networks benefit from SRDL resulting in enhanced deployment performance on both coarse-grained object categorisation tasks (CIFAR10, CIFAR100, Tiny ImageNet, and ImageNet) and fine-grained person instance identification tasks (Market-1501).

28.Localisation via Deep Imagination: learn the features not the map pdf

How many times does a human have to drive through the same area to become familiar with it? To begin with, we might first build a mental model of our surroundings. Upon revisiting this area, we can use this model to extrapolate to new unseen locations and imagine their appearance. Based on this, we propose an approach where an agent is capable of modelling new environments after a single visitation. To this end, we introduce "Deep Imagination", a combination of classical Visual-based Monte Carlo Localisation and deep learning. By making use of a feature embedded 3D map, the system can "imagine" the view from any novel location. These "imagined" views are contrasted with the current observation in order to estimate the agent's current location. In order to build the embedded map, we train a deep Siamese Fully Convolutional U-Net to perform dense feature extraction. By training these features to be generic, no additional training or fine tuning is required to adapt to new environments. Our results demonstrate the generality and transfer capability of our learnt dense features by training and evaluating on multiple datasets. Additionally, we include several visualizations of the feature representations and resulting 3D maps, as well as their application to localisation.

29.Fine-grained Classification using Heterogeneous Web Data and Auxiliary Categories pdf

Fine-grained classification remains a very challenging problem, because of the absence of well-labeled training data caused by the high cost of annotating a large number of fine-grained categories. In the extreme case, given a set of test categories without any well-labeled training data, the majority of existing works can be grouped into the following two research directions: 1) crawl noisy labeled web data for the test categories as training data, which is dubbed as webly supervised learning; 2) transfer the knowledge from auxiliary categories with well-labeled training data to the test categories, which corresponds to zero-shot learning setting. Nevertheless, the above two research directions still have critical issues to be addressed. For the first direction, web data have noisy labels and considerably different data distribution from test data. For the second direction, zero-shot learning is struggling to achieve compelling results compared with conventional supervised learning. The issues of the above two directions motivate us to develop a novel approach which can jointly exploit both noisy web training data from test categories and well-labeled training data from auxiliary categories. In particular, on one hand, we crawl web data for test categories as noisy training data. On the other hand, we transfer the knowledge from auxiliary categories with well-labeled training data to test categories by virtue of free semantic information (e.g., word vector) of all categories. Moreover, given the fact that web data are generally associated with additional textual information (e.g., title and tag), we extend our method by using the surrounding textual information of web data as privileged information. Extensive experiments show the effectiveness of our proposed methods.

30.iQIYI-VID: A Large Dataset for Multi-modal Person Identification pdf

Person identification in the wild is very challenging due to great variation in poses, face quality, clothes, makeup and so on. Traditional research, such as face recognition, person re-identification, and speaker recognition, often focuses on a single modal of information, which is inadequate to handle all the situations in practice. Multi-modal person identification is a more promising way that we can jointly utilize face, head, body, audio features, and so on. In this paper, we introduce iQIYI-VID, the largest video dataset for multi-modal person identification. It is composed of 600K video clips of 5,000 celebrities. These video clips are extracted from 400K hours of online videos of various types, ranging from movies, variety shows, TV series, to news broadcasting. All video clips pass through a careful human annotation process, and the error rate of labels is lower than 0.2%. We evaluated the state-of-art models of face recognition, person re-identification, and speaker recognition on the iQIYI-VID dataset. Experimental results show that these models are still far from being perfect for task of person identification in the wild. We further demonstrate that a simple fusion of multi-modal features can improve person identification considerably. We have released the dataset online to promote multi-modal person identification research.

31.CA3Net: Contextual-Attentional Attribute-Appearance Network for Person Re-Identification pdf

Person re-identification aims to identify the same pedestrian across non-overlapping camera views. Deep learning techniques have been applied for person re-identification recently, towards learning representation of pedestrian appearance. This paper presents a novel Contextual-Attentional Attribute-Appearance Network (CA3Net) for person re-identification. The CA3Net simultaneously exploits the complementarity between semantic attributes and visual appearance, the semantic context among attributes, visual attention on attributes as well as spatial dependencies among body parts, leading to discriminative and robust pedestrian representation. Specifically, an attribute network within CA3Net is designed with an Attention-LSTM module. It concentrates the network on latent image regions related to each attribute as well as exploits the semantic context among attributes by a LSTM module. An appearance network is developed to learn appearance features from the full body, horizontal and vertical body parts of pedestrians with spatial dependencies among body parts. The CA3Net jointly learns the attribute and appearance features in a multi-task learning manner, generating comprehensive representation of pedestrians. Extensive experiments on two challenging benchmarks, i.e., Market-1501 and DukeMTMC-reID datasets, have demonstrated the effectiveness of the proposed approach.

32.A Pretrained DenseNet Encoder for Brain Tumor Segmentation pdf

This article presents a convolutional neural network for the automatic segmentation of brain tumors in multimodal 3D MR images based on a U-net architecture.We evaluate the use of a densely connected convolutional network encoder (DenseNet) which was pretrained on the ImageNet data set. We detail two network architectures that can take into account multiple 3D images as inputs. This work aims to identify if a generic pretrained network can be used for very specific medical applications where the target data differ both in the number of spatial dimensions as well as in the number of inputs channels. Moreover in order to regularize this transfer learning task we only train the decoder part of the U-net architecture. We evaluate the effectiveness of the proposed approach on the BRATS 2018 segmentation challenge where we obtained dice scores of 0.79, 0.90, 0.85 and 95/% Hausdorff distance of 2.9mm, 3.95mm, and 6.48mm for enhanced tumor core, whole tumor and tumor core respectively on the validation set. This scores degrades to 0.77, 0.88, 0.78 and 95 /% Hausdorff distance of 3.6mm, 5.72mm, and 5.83mm on the testing set.

33.High Order Neural Networks for Video Classification pdf

Capturing spatiotemporal correlations is an essential topic in video classification. In this paper, we present high order operations as a generic family of building blocks for capturing high order correlations from high dimensional input video space. We prove that several successful architectures for visual classification tasks are in the family of high order neural networks, theoretical and experimental analysis demonstrates their underlying mechanism is high order. We also proposal a new LEarnable hiGh Order (LEGO) block, whose goal is to capture spatiotemporal correlation in a feedforward manner. Specifically, LEGO blocks implicitly learn the relation expressions for spatiotemporal features and use the learned relations to weight input features. This building block can be plugged into many neural network architectures, achieving evident improvement without introducing much overhead. On the task of video classification, even using RGB only without fine-tuning with other video datasets, our high order models can achieve results on par with or better than the existing state-of-the-art methods on both Something-Something (V1 and V2) and Charades datasets.

34.Unsupervised Learning in Reservoir Computing for EEG-based Emotion Recognition pdf

In real-world applications such as emotion recognition from recorded brain activity, data are captured from electrodes over time. These signals constitute a multidimensional time series. In this paper, Echo State Network (ESN), a recurrent neural network with a great success in time series prediction and classification, is optimized with different neural plasticity rules for classification of emotions based on electroencephalogram (EEG) time series. Actually, the neural plasticity rules are a kind of unsupervised learning adapted for the reservoir, i.e. the hidden layer of ESN. More specifically, an investigation of Oja's rule, BCM rule and gaussian intrinsic plasticity rule was carried out in the context of EEG-based emotion recognition. The study, also, includes a comparison of the offline and online training of the ESN. When testing on the well-known affective benchmark "DEAP dataset" which contains EEG signals from 32 subjects, we find that pretraining ESN with gaussian intrinsic plasticity enhanced the classification accuracy and outperformed the results achieved with an ESN pretrained with synaptic plasticity. Four classification problems were conducted in which the system complexity is increased and the discrimination is more challenging, i.e. inter-subject emotion discrimination. Our proposed method achieves higher performance over the state of the art methods.

35.Compressing Recurrent Neural Networks with Tensor Ring for Action Recognition pdf

Recurrent Neural Networks (RNNs) and their variants, such as Long-Short Term Memory (LSTM) networks, and Gated Recurrent Unit (GRU) networks, have achieved promising performance in sequential data modeling. The hidden layers in RNNs can be regarded as the memory units, which are helpful in storing information in sequential contexts. However, when dealing with high dimensional input data, such as video and text, the input-to-hidden linear transformation in RNNs brings high memory usage and huge computational cost. This makes the training of RNNs unscalable and difficult. To address this challenge, we propose a novel compact LSTM model, named as TR-LSTM, by utilizing the low-rank tensor ring decomposition (TRD) to reformulate the input-to-hidden transformation. Compared with other tensor decomposition methods, TR-LSTM is more stable. In addition, TR-LSTM can complete an end-to-end training and also provide a fundamental building block for RNNs in handling large input data. Experiments on real-world action recognition datasets have demonstrated the promising performance of the proposed TR-LSTM compared with the tensor train LSTM and other state-of-the-art competitors.

36.Fast Efficient Object Detection Using Selective Attention pdf

Deep learning object detectors achieve state-of-the-art accuracy at the expense of high computational overheads, impeding their utilization on embedded systems such as drones. A primary source of these overheads is the exhaustive classification of typically 10^4-10^5 regions per image. Given that most of these regions contain uninformative background, the detector designs seem extremely superfluous and inefficient. In contrast, biological vision systems leverage selective attention for fast and efficient object detection. Recent neuroscientific findings shedding new light on the mechanism behind selective attention allowed us to formulate a new hypothesis of object detection efficiency and subsequently introduce a new object detection paradigm. To that end, we leverage this knowledge to design a novel region proposal network and empirically show that it achieves high object detection performance on the COCO dataset. Moreover, the model uses two to three orders of magnitude fewer computations than state-of-the-art models and consequently achieves inference speeds exceeding 500 frames/s, thereby making it possible to achieve object detection on embedded systems.

37.Robust Visual Tracking using Multi-Frame Multi-Feature Joint Modeling pdf

It remains a huge challenge to design effective and efficient trackers under complex scenarios, including occlusions, illumination changes and pose variations. To cope with this problem, a promising solution is to integrate the temporal consistency across consecutive frames and multiple feature cues in a unified model. Motivated by this idea, we propose a novel correlation filter-based tracker in this work, in which the temporal relatedness is reconciled under a multi-task learning framework and the multiple feature cues are modeled using a multi-view learning approach. We demonstrate the resulting regression model can be efficiently learned by exploiting the structure of blockwise diagonal matrix. A fast blockwise diagonal matrix inversion algorithm is developed thereafter for efficient online tracking. Meanwhile, we incorporate an adaptive scale estimation mechanism to strengthen the stability of scale variation tracking. We implement our tracker using two types of features and test it on two benchmark datasets. Experimental results demonstrate the superiority of our proposed approach when compared with other state-of-the-art trackers. project homepage this http URL

38.FotonNet: A HW-Efficient Object Detection System Using 3D-Depth Segmentation and 2D-DNN Classifier pdf

Object detection and classification is one of the most important computer vision problems. Ever since the introduction of deep learning \cite{krizhevsky2012imagenet}, we have witnessed a dramatic increase in the accuracy of this object detection problem. However, most of these improvements have occurred using conventional 2D image processing. Recently, low-cost 3D-image sensors, such as the Microsoft Kinect (Time-of-Flight) or the Apple FaceID (Structured-Light), can provide 3D-depth or point cloud data that can be added to a convolutional neural network, acting as an extra set of dimensions. In our proposed approach, we introduce a new 2D + 3D system that takes the 3D-data to determine the object region followed by any conventional 2D-DNN, such as AlexNet. In this method, our approach can easily dissociate the information collection from the Point Cloud and 2D-Image data and combine both operations later. Hence, our system can use any existing trained 2D network on a large image dataset, and does not require a large 3D-depth dataset for new training. Experimental object detection results across 30 images show an accuracy of 0.67, versus 0.54 and 0.51 for RCNN and YOLO, respectively.

39.DeepSeeNet: A deep learning model for automated classification of patient-based age-related macular degeneration severity from color fundus photographs pdf

In assessing the severity of age-related macular degeneration (AMD), the Age-Related Eye Disease Study (AREDS) Simplified Severity Scale predicts the risk of progression to late AMD. However, its manual use requires the time-consuming participation of expert practitioners. While several automated deep learning (DL) systems have been developed for classifying color fundus photographs of individual eyes by AREDS severity score, none to date has utilized a patient-based scoring system that employs images from both eyes to assign a severity score. DeepSeeNet, a DL model, was developed to classify patients automatically by the AREDS Simplified Severity Scale (score 0-5) using bilateral color fundus images. DeepSeeNet was trained on 58,402 and tested on 900 images from the longitudinal follow up of 4,549 participants from AREDS. Gold standard labels were obtained using reading center grades. DeepSeeNet simulates the human grading process by first detecting individual AMD risk factors (drusen size; pigmentary abnormalities) for each eye and then calculating a patient-based AMD severity score using the AREDS Simplified Severity Scale. DeepSeeNet performed better on patient-based, multi-class classification (accuracy=0.671; kappa=0.558) than retinal specialists (accuracy=0.599; kappa=0.467) with high AUCs in the detection of large drusen (0.94), pigmentary abnormalities (0.93) and late AMD (0.97), respectively. DeepSeeNet demonstrated high accuracy with increased transparency in the automated assignment of individual patients to AMD risk categories based on the AREDS Simplified Severity Scale. These results highlight the potential of deep learning systems to assist and enhance clinical decision-making processes in AMD patients such as early AMD detection and risk prediction for developing late AMD. DeepSeeNet is publicly available on this https URL.

40.A Self-Adaptive Network For Multiple Sclerosis Lesion Segmentation From Multi-Contrast MRI With Various Imaging Protocols pdf

Deep neural networks (DNN) have shown promises in the lesion segmentation of multiple sclerosis (MS) from multicontrast MRI including T1, T2, proton density (PD) and FLAIR sequences. However, one challenge in deploying such networks into clinical practice is the variability of imaging protocols, which often differ from the training dataset as certain MRI sequences may be unavailable or unusable. Therefore, trained networks need to adapt to practical situations when imaging protocols are different in deployment. In this paper, we propose a DNN-based MS lesion segmentation framework with a novel technique called sequence dropout which can adapt to various combinations of input MRI sequences during deployment and achieve the maximal possible performance from the given input. In addition, with this framework, we studied the quantitative impact of each MRI sequence on the MS lesion segmentation task without training separate networks. Experiments were performed using the IEEE ISBI 2015 Longitudinal MS Lesion Challenge dataset and our method is currently ranked 2nd with a Dice similarity coefficient of 0.684. Furthermore, we showed our network achieved the maximal possible performance when one sequence is unavailable during deployment by comparing with separate networks trained on the corresponding input MRI sequences. In particular, we discovered T1 and PD have minor impact on segmentation performance while FLAIR is the predominant sequence. Experiments with multiple missing sequences were also performed and showed the robustness of our network.

41.Quantifying Human Behavior on the Block Design Test Through Automated Multi-Level Analysis of Overhead Video pdf

The block design test is a standardized, widely used neuropsychological assessment of visuospatial reasoning that involves a person recreating a series of given designs out of a set of colored blocks. In current testing procedures, an expert neuropsychologist observes a person's accuracy and completion time as well as overall impressions of the person's problem-solving procedures, errors, etc., thus obtaining a holistic though subjective and often qualitative view of the person's cognitive processes. We propose a new framework that combines room sensors and AI techniques to augment the information available to neuropsychologists from block design and similar tabletop assessments. In particular, a ceiling-mounted camera captures an overhead view of the table surface. From this video, we demonstrate how automated classification using machine learning can produce a frame-level description of the state of the block task and the person's actions over the course of each test problem. We also show how a sequence-comparison algorithm can classify one individual's problem-solving strategy relative to a database of simulated strategies, and how these quantitative results can be visualized for use by neuropsychologists.

42.Re-Identification with Consistent Attentive Siamese Networks pdf

We propose a new deep architecture for person re-identification (re-id). While re-id has seen much recent progress, spatial localization and view-invariant representation learning for robust cross-view matching remain key, unsolved problems. We address these questions by means of a new attention-driven Siamese learning architecture, called the Consistent Attentive Siamese Network. Our key innovations compared to existing, competing methods include (a) a flexible framework design that produces attention with only identity labels as supervision, (b) explicit mechanisms to enforce attention consistency among images of the same person, and (c) a new Siamese framework that integrates attention and attention consistency, producing principled supervisory signals as well as the first mechanism that can explain the reasoning behind the Siamese framework's predictions. We conduct extensive evaluations on the CUHK03-NP, DukeMTMC-ReID, and Market-1501 datasets, and establish a new state of the art, with our proposed method resulting in mAP performance improvements of 6.4%, 4.2%, and 1.2% respectively.

43.Visual-Texual Emotion Analysis with Deep Coupled Video and Danmu Neural Networks pdf

User emotion analysis toward videos is to automatically recognize the general emotional status of viewers from the multimedia content embedded in the online video stream. Existing works fall in two categories: 1) visual-based methods, which focus on visual content and extract a specific set of features of videos. However, it is generally hard to learn a mapping function from low-level video pixels to high-level emotion space due to great intra-class variance. 2) textual-based methods, which focus on the investigation of user-generated comments associated with videos. The learned word representations by traditional linguistic approaches typically lack emotion information and the global comments usually reflect viewers' high-level understandings rather than instantaneous emotions. To address these limitations, in this paper, we propose to jointly utilize video content and user-generated texts simultaneously for emotion analysis. In particular, we introduce exploiting a new type of user-generated texts, i.e., "danmu", which are real-time comments floating on the video and contain rich information to convey viewers' emotional opinions. To enhance the emotion discriminativeness of words in textual feature extraction, we propose Emotional Word Embedding (EWE) to learn text representations by jointly considering their semantics and emotions. Afterwards, we propose a novel visual-textual emotion analysis model with Deep Coupled Video and Danmu Neural networks (DCVDN), in which visual and textual features are synchronously extracted and fused to form a comprehensive representation by deep-canonically-correlated-autoencoder-based multi-view learning. Through extensive experiments on a self-crawled real-world video-danmu dataset, we prove that DCVDN significantly outperforms the state-of-the-art baselines.

44.Reducing Visual Confusion with Discriminative Attention pdf

Recent developments in gradient-based attention modeling have led to improved model interpretability by means of class-specific attention maps. A key limitation, however, of these approaches is that the resulting attention maps, while being well localized, are not class discriminative. In this paper, we address this limitation with a new learning framework that makes class-discriminative attention and cross-layer attention consistency a principled and explicit part of the learning process. Furthermore, our framework provides attention guidance to the model in an end-to-end fashion, resulting in better discriminability and reduced visual confusion. We conduct extensive experiments on various image classification benchmarks with our proposed framework and demonstrate its efficacy by means of improved classification accuracy including CIFAR-100 (+3.46%), Caltech-256 (+1.64%), ImageNet (+0.92%), CUB-200-2011 (+4.8%) and PASCAL VOC2012 (+5.78%).

45.Show, Attend and Translate: Unpaired Multi-Domain Image-to-Image Translation with Visual Attention pdf

Recently unpaired multi-domain image-to-image translation has attracted great interests and obtained remarkable progress, where a label vector is utilized to indicate multi-domain information. In this paper, we propose SAT (Show, Attend and Translate), an unified and explainable generative adversarial network equipped with visual attention that can perform unpaired image-to-image translation for multiple domains. By introducing an action vector, we treat the original translation tasks as problems of arithmetic addition and subtraction. Visual attention is applied to guarantee that only the regions relevant to the target domains are translated. Extensive experiments on a facial attribute dataset demonstrate the superiority of our approach and the generated attention masks better explain what SAT attends when translating images.

46.Global and Local Sensitivity Guided Key Salient Object Re-augmentation for Video Saliency Detection pdf

The existing still-static deep learning based saliency researches do not consider the weighting and highlighting of extracted features from different layers, all features contribute equally to the final saliency decision-making. Such methods always evenly detect all "potentially significant regions" and unable to highlight the key salient object, resulting in detection failure of dynamic scenes. In this paper, based on the fact that salient areas in videos are relatively small and concentrated, we propose a \textbf{key salient object re-augmentation method (KSORA) using top-down semantic knowledge and bottom-up feature guidance} to improve detection accuracy in video scenes. KSORA includes two sub-modules (WFE and KOS): WFE processes local salient feature selection using bottom-up strategy, while KOS ranks each object in global fashion by top-down statistical knowledge, and chooses the most critical object area for local enhancement. The proposed KSORA can not only strengthen the saliency value of the local key salient object but also ensure global saliency consistency. Results on three benchmark datasets suggest that our model has the capability of improving the detection accuracy on complex scenes. The significant performance of KSORA, with a speed of 17FPS on modern GPUs, has been verified by comparisons with other ten state-of-the-art algorithms.

47.Multi-scale 3D Convolution Network for Video Based Person Re-Identification pdf

This paper proposes a two-stream convolution network to extract spatial and temporal cues for video based person Re-Identification (ReID). A temporal stream in this network is constructed by inserting several Multi-scale 3D (M3D) convolution layers into a 2D CNN network. The resulting M3D convolution network introduces a fraction of parameters into the 2D CNN, but gains the ability of multi-scale temporal feature learning. With this compact architecture, M3D convolution network is also more efficient and easier to optimize than existing 3D convolution networks. The temporal stream further involves Residual Attention Layers (RAL) to refine the temporal features. By jointly learning spatial-temporal attention masks in a residual manner, RAL identifies the discriminative spatial regions and temporal cues. The other stream in our network is implemented with a 2D CNN for spatial feature extraction. The spatial and temporal features from two streams are finally fused for the video based person ReID. Evaluations on three widely used benchmarks datasets, i.e., MARS, PRID2011, and iLIDS-VID demonstrate the substantial advantages of our method over existing 3D convolution networks and state-of-art methods.

48.Indoor GeoNet: Weakly Supervised Hybrid Learning for Depth and Pose Estimation pdf

Humans naturally perceive a 3D scene in front of them through accumulation of information obtained from multiple interconnected projections of the scene and by interpreting their correspondence. This phenomenon has inspired artificial intelligence models to extract the depth and view angle of the observed scene by modeling the correspondence between different views of that scene. Our paper is built upon previous works in the field of unsupervised depth and relative camera pose estimation from temporal consecutive video frames using deep learning (DL) models. Our approach uses a hybrid learning framework introduced in a recent work called GeoNet, which leverages geometric constraints in the 3D scenes to synthesize a novel view from intermediate DL-based predicted depth and relative pose. However, the state-of-the-art unsupervised depth and pose estimation DL models are exclusively trained/tested on a few available outdoor scene datasets and we have shown they are hardly transferable to new scenes, especially from indoor environments, in which estimation requires higher precision and dealing with probable occlusions. This paper introduces "Indoor GeoNet", a weakly supervised depth and camera pose estimation model targeted for indoor scenes. In Indoor GeoNet, we take advantage of the availability of indoor RGBD datasets collected by human or robot navigators, and added partial (i.e. weak) supervision in depth training into the model. Experimental results showed that our model effectively generalizes to new scenes from different buildings. Indoor GeoNet demonstrated significant depth and pose estimation error reduction when compared to the original GeoNet, while showing 3 times more reconstruction accuracy in synthesizing novel views in indoor environments.

49.Segregated Temporal Assembly Recurrent Networks for Weakly Supervised Multiple Action Detection pdf

This paper proposes a segregated temporal assembly recurrent (STAR) network for weakly-supervised multiple action detection. The model learns from untrimmed videos with only supervision of video-level labels and makes prediction of intervals of multiple actions. Specifically, we first assemble video clips according to class labels by an attention mechanism that learns class-variable attention weights and thus helps the noise relieving from background or other actions. Secondly, we build temporal relationship between actions by feeding the assembled features into an enhanced recurrent neural network. Finally, we transform the output of recurrent neural network into the corresponding action distribution. In order to generate more precise temporal proposals, we design a score term called segregated temporal gradient-weighted class activation mapping (ST-GradCAM) fused with attention weights. Experiments on THUMOS'14 and ActivityNet1.3 datasets show that our approach outperforms the state-of-the-art weakly-supervised method, and performs at par with the fully-supervised counterparts.

50.An Efficient Transfer Learning Technique by Using Final Fully-Connected Layer Output Features of Deep Networks pdf

In this paper, we propose a computationally efficient transfer learning approach using the output vector of final fully-connected layer of deep convolutional neural networks for classification. Our proposed technique uses a single layer perceptron classifier designed with hyper-parameters to focus on improving computational efficiency without adversely affecting the performance of classification compared to the baseline technique. Our investigations show that our technique converges much faster than baseline yielding very competitive classification results. We execute thorough experiments to understand the impact of similarity between pre-trained and new classes, similarity among new classes, number of training samples in the performance of classification using transfer learning of the final fully-connected layer's output features.

51.Unsupervised Domain Adaptation: An Adaptive Feature Norm Approach pdf

Unsupervised domain adaptation aims to mitigate the domain shift when transferring knowledge from a supervised source domain to an unsupervised target domain. Adversarial Feature Alignment has been successfully explored to minimize the domain discrepancy. However, existing methods are usually struggling to optimize mixed learning objectives and vulnerable to negative transfer when two domains do not share the identical label space. In this paper, we empirically reveal that the erratic discrimination of target domain mainly reflects in its much lower feature norm value with respect to that of the source domain. We present a non-parametric Adaptive Feature Norm AFN approach, which is independent of the association between label spaces of the two domains. We demonstrate that adapting feature norms of source and target domains to achieve equilibrium over a large range of values can result in significant domain transfer gains. Without bells and whistles but a few lines of code, our method largely lifts the discrimination of target domain (23.7% from the Source Only in VisDA2017) and achieves the new state of the art under the vanilla setting. Furthermore, as our approach does not require to deliberately align the feature distributions, it is robust to negative transfer and can outperform the existing approaches under the partial setting by an extremely large margin (9.8% on Office-Home and 14.1% on VisDA2017). Code is available at this https URL. We are responsible for the reproducibility of our method.

52.Predictive and Semantic Layout Estimation for Robotic Applications in Manhattan Worlds pdf

This paper describes an approach to automatically extracting floor plans from the kinds of incomplete measurements that could be acquired by an autonomous mobile robot. The approach proceeds by reasoning about extended structural layout surfaces which are automatically extracted from the available data. The scheme can be run in an online manner to build water tight representations of the environment. The system effectively speculates about room boundaries and free space regions which provides useful guidance to subsequent motion planning systems. Experimental results are presented on multiple data sets.

53.Pixel-Anchor: A Fast Oriented Scene Text Detector with Combined Networks pdf

Recently, semantic segmentation and general object detection frameworks have been widely adopted by scene text detecting tasks. However, both of them alone have obvious shortcomings in practice. In this paper, we propose a novel end-to-end trainable deep neural network framework, named Pixel-Anchor, which combines semantic segmentation and SSD in one network by feature sharing and anchor-level attention mechanism to detect oriented scene text. To deal with scene text which has large variances in size and aspect ratio, we combine FPN and ASPP operation as our encoder-decoder structure in the semantic segmentation part, and propose a novel Adaptive Predictor Layer in the SSD. Pixel-Anchor detects scene text in a single network forward pass, no complex post-processing other than an efficient fusion Non-Maximum Suppression is involved. We have benchmarked the proposed Pixel-Anchor on the public datasets. Pixel-Anchor outperforms the competing methods in terms of text localization accuracy and run speed, more specifically, on the ICDAR 2015 dataset, the proposed algorithm achieves an F-score of 0.8768 at 10 FPS for 960 x 1728 resolution images.

54.Multimodal Densenet pdf

Humans make accurate decisions by interpreting complex data from multiple sources. Medical diagnostics, in particular, often hinge on human interpretation of multi-modal information. In order for artificial intelligence to make progress in automated, objective, and accurate diagnosis and prognosis, methods to fuse information from multiple medical imaging modalities are required. However, combining information from multiple data sources has several challenges, as current deep learning architectures lack the ability to extract useful representations from multimodal information, and often simple concatenation is used to fuse such information. In this work, we propose Multimodal DenseNet, a novel architecture for fusing multimodal data. Instead of focusing on concatenation or early and late fusion, our proposed architectures fuses information over several layers and gives the model flexibility in how it combines information from multiple sources. We apply this architecture to the challenge of polyp characterization and landmark identification in endoscopy. Features from white light images are fused with features from narrow band imaging or depth maps. This study demonstrates that Multimodal DenseNet outperforms monomodal classification as well as other multimodal fusion techniques by a significant margin on two different datasets.

55.Facial Expression and Peripheral Physiology Fusion to Decode Individualized Affective Experience pdf

In this paper, we present a multimodal approach to simultaneously analyze facial movements and several peripheral physiological signals to decode individualized affective experiences under positive and negative emotional contexts, while considering their personalized resting dynamics. We propose a person-specific recurrence network to quantify the dynamics present in the person's facial movements and physiological data. Facial movement is represented using a robust head vs. 3D face landmark localization and tracking approach, and physiological data are processed by extracting known attributes related to the underlying affective experience. The dynamical coupling between different input modalities is then assessed through the extraction of several complex recurrent network metrics. Inference models are then trained using these metrics as features to predict individual's affective experience in a given context, after their resting dynamics are excluded from their response. We validated our approach using a multimodal dataset consists of (i) facial videos and (ii) several peripheral physiological signals, synchronously recorded from 12 participants while watching 4 emotion-eliciting video-based stimuli. The affective experience prediction results signified that our multimodal fusion method improves the prediction accuracy up to 19% when compared to the prediction using only one or a subset of the input modalities. Furthermore, we gained prediction improvement for affective experience by considering the effect of individualized resting dynamics.

56.Temporal Recurrent Networks for Online Action Detection pdf

Most work on temporal action detection is formulated in an offline manner, in which the start and end times of actions are determined after the entire video is fully observed. However, real-time applications including surveillance and driver assistance systems require identifying actions as soon as each video frame arrives, based only on current and historical observations. In this paper, we propose a novel framework, Temporal Recurrent Networks (TRNs), to model greater temporal context of a video frame by simultaneously performing online action detection and anticipation of the immediate future. At each moment in time, our approach makes use of both accumulated historical evidence and predicted future information to better recognize the action that is currently occurring, and integrates both of these into a unified end-to-end architecture. We evaluate our approach on two popular online action detection datasets, HDD and TVSeries, as well as another widely used dataset, THUMOS'14. The results show that TRN significantly outperforms the state-of-the-art.

57.Deep Siamese Networks with Bayesian non-Parametrics for Video Object Tracking pdf

We present a novel algorithm utilizing a deep Siamese neural network as a general object similarity function in combination with a Bayesian optimization (BO) framework to encode spatio-temporal information for efficient object tracking in video. In particular, we treat the video tracking problem as a dynamic (i.e. temporally-evolving) optimization problem. Using Gaussian Process priors, we model a dynamic objective function representing the location of a tracked object in each frame. By exploiting temporal correlations, the proposed method queries the search space in a statistically principled and efficient way, offering several benefits over current state of the art video tracking methods.

58.RGB-based 3D Hand Pose Estimation via Privileged Learning with Depth Images pdf

This paper proposes a method for hand pose estimation from RGB images that uses both external large-scale depth image datasets and paired depth and RGB images as privileged information at training time. We show that providing depth information during training significantly improves performance of pose estimation from RGB images during testing. We explore different ways of using this privileged information: (1) using depth data to initially train a depth-based network, (2) using the features from the depth-based network of the paired depth images to constrain mid-level RGB network weights, and (3) using the foreground mask, obtained from the depth data, to suppress the responses from the background area. By using paired RGB and depth images, we are able to supervise the RGB-based network to learn middle layer features that mimic that of the corresponding depth-based network, which is trained on large-scale, accurately annotated depth data. During testing, when only an RGB image is available, our method produces accurate 3D hand pose predictions. Our method is also tested on 2D hand pose estimation. Experiments on three public datasets show that the method outperforms the state-of-the-art methods for hand pose estimation using RGB image input.

59.Transfer Learning with Deep CNNs for Gender Recognition and Age Estimation pdf

In this project, competition-winning deep neural networks with pretrained weights are used for image-based gender recognition and age estimation. Transfer learning is explored using both VGG19 and VGGFace pretrained models by testing the effects of changes in various design schemes and training parameters in order to improve prediction accuracy. Training techniques such as input standardization, data augmentation, and label distribution age encoding are compared. Finally, a hierarchy of deep CNNs is tested that first classifies subjects by gender, and then uses separate male and female age models to predict age. A gender recognition accuracy of 98.7% and an MAE of 4.1 years is achieved. This paper shows that, with proper training techniques, good results can be obtained by retasking existing convolutional filters towards a new purpose.

60.Implementation of Robust Face Recognition System Using Live Video Feed Based on CNN pdf

The way to accurately and effectively identify people has always been an interesting topic in research and industry. With the rapid development of artificial intelligence in recent years, facial recognition gains lots of attention due to prompting the development of emerging identification methods. Compared to traditional card recognition, fingerprint recognition and iris recognition, face recognition has many advantages including non-contact interface, high concurrency, and user-friendly usage. It has high potential to be used in government, public facilities, security, e-commerce, retailing, education and many other fields. With the development of deep learning and the introduction of deep convolutional neural networks, the accuracy and speed of face recognition have made great strides. However, the results from different networks and models are very different with different system architecture. Furthermore, it could take significant amount of data storage space and data processing time for the face recognition system with video feed, if the system stores images and features of human faces. In this paper, facial features are extracted by merging and comparing multiple models, and then a deep neural network is constructed to train and construct the combined features. In this way, the advantages of multiple models can be combined to mention the recognition accuracy. After getting a model with high accuracy, we build a product model. The model will take a human face image and extract it into a vector. Then the distance between vectors are compared to determine if two faces on different picture belongs to the same person. The proposed approach reduces data storage space and data processing time for the face recognition system with video feed scientifically with our proposed system architecture.

61.On Matching Faces with Alterations due to Plastic Surgery and Disguise pdf

Plastic surgery and disguise variations are two of the most challenging co-variates of face recognition. The state-of-art deep learning models are not sufficiently successful due to the availability of limited training samples. In this paper, a novel framework is proposed which transfers fundamental visual features learnt from a generic image dataset to supplement a supervised face recognition model. The proposed algorithm combines off-the-shelf supervised classifier and a generic, task independent network which encodes information related to basic visual cues such as color, shape, and texture. Experiments are performed on IIITD plastic surgery face dataset and Disguised Faces in the Wild (DFW) dataset. Results showcase that the proposed algorithm achieves state of the art results on both the datasets. Specifically on the DFW database, the proposed algorithm yields over 87% verification accuracy at 1% false accept rate which is 53.8% better than baseline results computed using VGGFace.

62.Learning to infer: RL-based search for DNN primitive selection on Heterogeneous Embedded Systems pdf

Deep Learning is increasingly being adopted by industry for computer vision applications running on embedded devices. While Convolutional Neural Networks' accuracy has achieved a mature and remarkable state, inference latency and throughput are a major concern especially when targeting low-cost and low-power embedded platforms. CNNs' inference latency may become a bottleneck for Deep Learning adoption by industry, as it is a crucial specification for many real-time processes. Furthermore, deployment of CNNs across heterogeneous platforms presents major compatibility issues due to vendor-specific technology and acceleration libraries. In this work, we present QS-DNN, a fully automatic search based on Reinforcement Learning which, combined with an inference engine optimizer, efficiently explores through the design space and empirically finds the optimal combinations of libraries and primitives to speed up the inference of CNNs on heterogeneous embedded devices. We show that, an optimized combination can achieve 45x speedup in inference latency on CPU compared to a dependency-free baseline and 2x on average on GPGPU compared to the best vendor library. Further, we demonstrate that, the quality of results and time "to-solution" is much better than with Random Search and achieves up to 15x better results for a short-time search.

63.Image-to-GPS Verification Through A Bottom-Up Pattern Matching Network pdf

The image-to-GPS verification problem asks whether a given image is taken at a claimed GPS location. In this paper, we treat it as an image verification problem -- whether a query image is taken at the same place as a reference image retrieved at the claimed GPS location. We make three major contributions: 1) we propose a novel custom bottom-up pattern matching (BUPM) deep neural network solution; 2) we demonstrate that the verification can be directly done by cross-checking a perspective-looking query image and a panorama reference image, and 3) we collect and clean a dataset of 30K pairs query and reference. Our experimental results show that the proposed BUPM solution outperforms the state-of-the-art solutions in terms of both verification and localization.

64.RePr: Improved Training of Convolutional Filters pdf

A well-trained Convolutional Neural Network can easily be pruned without significant loss of performance. This is because of unnecessary overlap in the features captured by the network's filters. Innovations in network architecture such as skip/dense connections and Inception units have mitigated this problem to some extent, but these improvements come with increased computation and memory requirements at run-time. We attempt to address this problem from another angle - not by changing the network structure but by altering the training method. We show that by temporarily pruning and then restoring a subset of the model's filters, and repeating this process cyclically, overlap in the learned features is reduced, producing improved generalization. We show that the existing model-pruning criteria are not optimal for selecting filters to prune in this context and introduce inter-filter orthogonality as the ranking criteria to determine under-expressive filters. Our method is applicable both to vanilla convolutional networks and more complex modern architectures, and improves the performance across a variety of tasks, especially when applied to smaller networks.

65.CIFAR10 to Compare Visual Recognition Performance between Deep Neural Networks and Humans pdf

Visual object recognition plays an essential role in human daily life. This ability is so efficient that we can recognize a face or an object seemingly without effort, though they may vary in position, scale, pose, and illumination. In the field of computer vision, a large number of studies have been carried out to build a human-like object recognition system. Recently, deep neural networks have shown impressive progress in object classification performance, and have been reported to surpass humans. Yet there is still lack of thorough and fair comparison between humans and artificial recognition systems. While some studies consider artificially degraded images, human recognition performance on dataset widely used for deep neural networks has not been fully evaluated. The present paper carries out an extensive experiment to evaluate human classification accuracy on CIFAR10, a well-known dataset of natural images. This then allows for a fair comparison with the state-of-the-art deep neural networks. Our CIFAR10-based evaluations show very efficient object recognition of recent CNNs but, at the same time, prove that they are still far from human-level capability of generalization. Moreover, a detailed investigation using multiple levels of difficulty reveals that easy images for humans may not be easy for deep neural networks. Such images form a subset of CIFAR10 that can be employed to evaluate and improve future neural networks.

66.Deep Learning with Inaccurate Training Data for Image Restoration pdf

In many applications of deep learning, particularly those in image restoration, it is either very difficult, prohibitively expensive, or outright impossible to obtain paired training data precisely as in the real world. In such cases, one is forced to use synthesized paired data to train the deep convolutional neural network (DCNN). However, due to the unavoidable generalization error in statistical learning, the synthetically trained DCNN often performs poorly on real world data. To overcome this problem, we propose a new general training method that can compensate for, to a large extent, the generalization errors of synthetically trained DCNNs.

67.DeepConsensus: using the consensus of features from multiple layers to attain robust image classification pdf

We consider a classifier whose test set is exposed to various perturbations that are not present in the training set. These test samples still contain enough features to map them to the same class as their unperturbed counterpart. Current architectures exhibit rapid degradation of accuracy when trained on standard datasets but then used to classify perturbed samples of that data. To address this, we present a novel architecture named DeepConsensus that significantly improves generalization to these test-time perturbations. Our key insight is that deep neural networks should directly consider summaries of low and high level features when making classifications. Existing convolutional neural networks can be augmented with DeepConsensus, leading to improved resistance against large and small perturbations on MNIST, EMNIST, FashionMNIST, CIFAR10 and SVHN datasets.

68.GLStyleNet: Higher Quality Style Transfer Combining Global and Local Pyramid Features pdf

Recent studies using deep neural networks have shown remarkable success in style transfer especially for artistic and photo-realistic images. However, the approaches using global feature correlations fail to capture small, intricate textures and maintain correct texture scales of the artworks, and the approaches based on local patches are defective on global effect. In this paper, we present a novel feature pyramid fusion neural network, dubbed GLStyleNet, which sufficiently takes into consideration multi-scale and multi-level pyramid features by best aggregating layers across a VGG network, and performs style transfer hierarchically with multiple losses of different scales. Our proposed method retains high-frequency pixel information and low frequency construct information of images from two aspects: loss function constraint and feature fusion. Our approach is not only flexible to adjust the trade-off between content and style, but also controllable between global and local. Compared to state-of-the-art methods, our method can transfer not just large-scale, obvious style cues but also subtle, exquisite ones, and dramatically improves the quality of style transfer. We demonstrate the effectiveness of our approach on portrait style transfer, artistic style transfer, photo-realistic style transfer and Chinese ancient painting style transfer tasks. Experimental results indicate that our unified approach improves image style transfer quality over previous state-of-the-art methods, while also accelerating the whole process in a certain extent. Our code is available at this https URL.

69.Exploit the Connectivity: Multi-Object Tracking with TrackletNet pdf

Multi-object tracking (MOT) is an important and practical task related to both surveillance systems and moving camera applications, such as autonomous driving and robotic vision. However, due to unreliable detection, occlusion and fast camera motion, tracked targets can be easily lost, which makes MOT very challenging. Most recent works treat tracking as a re-identification (Re-ID) task, but how to combine appearance and temporal features is still not well addressed. In this paper, we propose an innovative and effective tracking method called TrackletNet Tracker (TNT) that combines temporal and appearance information together as a unified framework. First, we define a graph model which treats each tracklet as a vertex. The tracklets are generated by appearance similarity with CNN features and intersection-over-union (IOU) with epipolar constraints to compensate camera movement between adjacent frames. Then, for every pair of two tracklets, the similarity is measured by our designed multi-scale TrackletNet. Afterwards, the tracklets are clustered into groups which represent individual object IDs. Our proposed TNT has the ability to handle most of the challenges in MOT, and achieve promising results on MOT16 and MOT17 benchmark datasets compared with other state-of-the-art methods.

70.Optical Flow Based Online Moving Foreground Analysis pdf

Obtained by moving object detection, the foreground mask result is unshaped and can not be directly used in most subsequent processes. In this paper, we focus on this problem and address it by constructing an optical flow based moving foreground analysis framework. During the processing procedure, the foreground masks are analyzed and segmented through two complementary clustering algorithms. As a result, we obtain the instance-level information like the number, location and size of moving objects. The experimental result show that our method adapts itself to the problem and performs well enough for practical applications.

71.Iris Presentation Attack Detection Based on Photometric Stereo Features pdf

We propose a new iris presentation attack detection method using three-dimensional features of an observed iris region estimated by photometric stereo. Our implementation uses a pair of iris images acquired by a common commercial iris sensor (LG 4000). No hardware modifications of any kind are required. Our approach should be applicable to any iris sensor that can illuminate the eye from two different directions. Each iris image in the pair is captured under near-infrared illumination at a different angle relative to the eye. Photometric stereo is used to estimate surface normal vectors in the non-occluded portions of the iris region. The variability of the normal vectors is used as the presentation attack detection score. This score is larger for a texture that is irregularly opaque and printed on a convex contact lens, and is smaller for an authentic iris texture. Thus the problem is formulated as binary classification into (a) an eye wearing textured contact lens and (b) the texture of an actual iris surface (possibly seen through a clear contact lens). Experiments were carried out on a database of approx. 2,900 iris image pairs acquired from approx. 100 subjects. Our method was able to correctly classify over 95% of samples when tested on contact lens brands unseen in training, and over 98% of samples when the contact lens brand was seen during training. The source codes of the method are made available to other researchers.

72.Matching RGB Images to CAD Models for Object Pose Estimation pdf

We propose a novel method for 3D object pose estimation in RGB images, which does not require pose annotations of objects in images in the training stage. We tackle the pose estimation problem by learning how to establish correspondences between RGB images and rendered depth images of CAD models. During training, our approach only requires textureless CAD models and aligned RGB-D frames of a subset of object instances, without explicitly requiring pose annotations for the RGB images. We employ a deep quadruplet convolutional neural network for joint learning of suitable keypoints and their associated descriptors in pairs of rendered depth images which can be matched across modalities with aligned RGB-D views. During testing, keypoints are extracted from a query RGB image and matched to keypoints extracted from rendered depth images, followed by establishing 2D-3D correspondences. The object's pose is then estimated using the RANSAC and PnP algorithms. We conduct experiments on the recently introduced Pix3D dataset and demonstrate the efficacy of our proposed approach in object pose estimation as well as generalization to object instances not seen during training.

73.PointConv: Deep Convolutional Networks on 3D Point Clouds pdf

Unlike images which are represented in regular dense grids, 3D point clouds are irregular and unordered, hence applying convolution on them can be difficult. In this paper, we extend the dynamic filter to a new convolution operation, named PointConv. PointConv can be applied on point clouds to build deep convolutional networks. We treat convolution kernels as nonlinear functions of the local coordinates of 3D points comprised of weight and density functions. With respect to a given point, the weight functions are learned with multi-layer perceptron networks and the density functions through kernel density estimation. A novel reformulation is proposed for efficiently computing the weight functions, which allowed us to dramatically scale up the network and significantly improve its performance. The learned convolution kernel can be used to compute translation-invariant and permutation-invariant convolution on any point set in the 3D space. Besides, PointConv can also be used as deconvolution operators to propagate features from a subsampled point cloud back to its original resolution. Experiments on ModelNet40, ShapeNet, and ScanNet show that deep convolutional neural networks built on PointConv are able to achieve state-of-the-art on challenging semantic segmentation benchmarks on 3D point clouds. Besides, our experiments converting CIFAR-10 into a point cloud showed that networks built on PointConv can match the performance of convolutional networks in 2D images of a similar structure.

74.GroundNet: Segmentation-Aware Monocular Ground Plane Estimation with Geometric Consistency pdf

We focus on the problem of estimating the orientation of the ground plane with respect to a mobile monocular camera platform (e.g., ground robot, wearable camera, assistive robotic platform). To address this problem, we formulate the ground plane estimation problem as an inter-mingled multi-task prediction problem by jointly optimizing for point-wise surface normal direction, 2D ground segmentation, and depth estimates. Our proposed model -- GroundNet -- estimates the ground normal in two streams separately and then a consistency loss is applied on top of the two streams to enforce geometric consistency. A semantic segmentation stream is used to isolate the ground regions and are used to selectively back-propagate parameter updates only through the ground regions in the image. Our experiments on KITTI and ApolloScape datasets verify that the GroundNet is able to predict consistent depth and normal within the ground region. It also achieves top performance on ground plane normal estimation and horizon line detection.

75.Open-vocabulary Phrase Detection pdf

Most existing work that grounds natural language phrases in images starts with the assumption that the phrase in question is relevant to the image. In this paper we address a more realistic version of the natural language grounding task where we must both identify whether the phrase is relevant to an image and localize the phrase. This can also be viewed as a generalization of object detection to an open-ended vocabulary, essentially introducing elements of few- and zero-shot detection. We propose a Phrase R-CNN network for this task that extends Faster R-CNN to relate image regions and phrases. By carefully initializing the classification layers of our network using canonical correlation analysis (CCA), we encourage a solution that is more discerning when reasoning between similar phrases, resulting in over double the performance compared to a naive adaptation on two popular phrase grounding datasets, Flickr30K Entities and ReferIt Game, with test-time phrase vocabulary sizes of 5K and 39K, respectively.

76.Sequential Image-based Attention Network for Inferring Force Estimation without Haptic Sensor pdf

Humans can infer approximate interaction force between objects from only vision information because we already have learned it through experiences. Based on this idea, we propose a recurrent convolutional neural network-based method using sequential images for inferring interaction force without using a haptic sensor. For training and validating deep learning methods, we collected a large number of images and corresponding interaction forces through an electronic motor-based device. To concentrate on changing shapes of a target object by the external force in images, we propose a sequential image-based attention module, which learns a salient model from temporal dynamics. The proposed sequential image-based attention module consists of a sequential spatial attention module and a sequential channel attention module, which are extended to exploit multiple sequential images. For gaining better accuracy, we also created a weighted average pooling layer for both spatial and channel attention modules. The extensive experimental results verified that the proposed method successfully infers interaction forces under the various conditions, such as different target materials, illumination changes, and external force directions.

77.Stacking-Based Deep Neural Network: Deep Analytic Network for Pattern Classification pdf

Stacking-based deep neural network (S-DNN) is aggregated with pluralities of basic learning modules, one after another, to synthesize a deep neural network (DNN) alternative for pattern classification. Contrary to the DNNs trained end to end by backpropagation (BP), each S-DNN layer, i.e., a self-learnable module, is to be trained decisively and independently without BP intervention. In this paper, a ridge regression-based S-DNN, dubbed deep analytic network (DAN), along with its kernelization (K-DAN), are devised for multilayer feature re-learning from the pre-extracted baseline features and the structured features. Our theoretical formulation demonstrates that DAN/K-DAN re-learn by perturbing the intra/inter-class variations, apart from diminishing the prediction errors. We scrutinize the DAN/K-DAN performance for pattern classification on datasets of varying domains - faces, handwritten digits, generic objects, to name a few. Unlike the typical BP-optimized DNNs to be trained from gigantic datasets by GPU, we disclose that DAN/K-DAN are trainable using only CPU even for small-scale training sets. Our experimental results disclose that DAN/K-DAN outperform the present S-DNNs and also the BP-trained DNNs, including multiplayer perceptron, deep belief network, etc., without data augmentation applied.

78.A Study of Human Body Characteristics Effect on Micro-Doppler-Based Person Identification using Deep Learning pdf

Obtaining a smart surveillance requires a sensing system that can capture accurate and detailed information for the human walking style. The radar micro-Doppler ($\boldsymbolμ$-D) analysis is proved to be a reliable metric for studying human locomotions. Thus, $\boldsymbolμ$-D signatures can be used to identify humans based on their walking styles. Additionally, the signatures contain information about the radar cross section (RCS) of the moving subject. This paper investigates the effect of human body characteristics on human identification based on their $\boldsymbolμ$-D signatures. In our proposed experimental setup, a treadmill is used to collect $\boldsymbolμ$-D signatures of 22 subjects with different genders and body characteristics. Convolutional autoencoders (CAE) are then used to extract the latent space representation from the $\boldsymbolμ$-D signatures. It is then interpreted in two dimensions using t-distributed stochastic neighbor embedding (t-SNE). Our study shows that the body mass index (BMI) has a correlation with the $\boldsymbolμ$-D signature of the walking subject. A 50-layer deep residual network is then trained to identify the walking subject based on the $\boldsymbolμ$-D signature. We achieve an accuracy of 98% on the test set with high signal-to-noise-ratio (SNR) and 84% in case of different SNR levels.

79.Optical Flow Dataset and Benchmark for Visual Crowd Analysis pdf

The performance of optical flow algorithms greatly depends on the specifics of the content and the application for which it is used. Existing and well established optical flow datasets are limited to rather particular contents from which none is close to crowd behavior analysis; whereas such applications heavily utilize optical flow. We introduce a new optical flow dataset exploiting the possibilities of a recent video engine to generate sequences with ground-truth optical flow for large crowds in different scenarios. We break with the development of the last decade of introducing ever increasing displacements to pose new difficulties. Instead we focus on real-world surveillance scenarios where numerous small, partly independent, non rigidly moving objects observed over a long temporal range pose a challenge. By evaluating different optical flow algorithms, we find that results of established datasets can not be transferred to these new challenges. In exhaustive experiments we are able to provide new insight into optical flow for crowd analysis. Finally, the results have been validated on the real-world UCF crowd tracking benchmark while achieving competitive results compared to more sophisticated state-of-the-art crowd tracking approaches.

80.Edge-Based Blur Kernel Estimation Using Sparse Representation and Self-Similarity pdf

Blind image deconvolution is the problem of recovering the latent image from the only observed blurry image when the blur kernel is unknown. In this paper, we propose an edge-based blur kernel estimation method for blind motion deconvolution. In our previous work, we incorporate both sparse representation and self-similarity of image patches as priors into our blind deconvolution model to regularize the recovery of the latent image. Since almost any natural image has properties of sparsity and multi-scale self-similarity, we construct a sparsity regularizer and a cross-scale non-local regularizer based on our patch priors. It has been observed that our regularizers often favor sharp images over blurry ones only for image patches of the salient edges and thus we define an edge mask to locate salient edges that we want to apply our regularizers. Experimental results on both simulated and real blurry images demonstrate that our method outperforms existing state-of-the-art blind deblurring methods even for handling of very large blurs, thanks to the use of the edge mask.

81.Recurrence to the Rescue: Towards Causal Spatiotemporal Representations pdf

Recently, three dimensional (3D) convolutional neural networks (CNNs) have emerged as dominant methods to capture spatiotemporal representations, by adding to pre-existing 2D CNNs a third, temporal dimension. Such 3D CNNs, however, are anti-causal (i.e., they exploit information from both the past and the future to produce feature representations, thus preventing their use in online settings), constrain the temporal reasoning horizon to the size of the temporal convolution kernel, and are not temporal resolution-preserving for video sequence-to-sequence modelling, as, e.g., in spatiotemporal action detection. To address these serious limitations, we present a new architecture for the causal/online spatiotemporal representation of videos. Namely, we propose a recurrent convolutional network (RCN), which relies on recurrence to capture the temporal context across frames at every level of network depth. Our network decomposes 3D convolutions into (1) a 2D spatial convolution component, and (2) an additional hidden state $1\times 1$ convolution applied across time. The hidden state at any time $t$ is assumed to depend on the hidden state at $t-1$ and on the current output of the spatial convolution component. As a result, the proposed network: (i) provides flexible temporal reasoning, (ii) produces causal outputs, and (iii) preserves temporal resolution. Our experiments on the large-scale large "Kinetics" dataset show that the proposed method achieves superior performance compared to 3D CNNs, while being causal and using fewer parameters.

82.Batch Feature Erasing for Person Re-identification and Beyond pdf

This paper presents a new training mechanism called Batch Feature Erasing (BFE) for person re-identification. We apply this strategy to train a novel network with two branches and employing the ResNet-50 as the backbone. The two branches consist of a conventional global branch and a feature erasing branch where the BFE strategy is applied. When training the feature erasing branch, we randomly erase the same region of all the feature maps in a batch. The network then concatenates features from the two branches for person re-identification. Albeit simple, our method achieves state-of-the-art on person re-identification and is applicable to general metric learning tasks in image retrieval problems. For instance, we achieve 75.4% Rank1 accuracy on the CUHK03-Detect dataset and 83.0% Recall-1 score on the Stanford Online Products dataset, outperforming the existed works by a large margin (more than 6%).

83.R2CNN++: Multi-Dimensional Attention Based Rotation Invariant Detector with Robust Anchor Strategy pdf

Object detection plays a vital role in natural scene and aerial scene and is full of challenges. Although many advanced algorithms have succeeded in the natural scene, the progress in the aerial scene has been slow due to the complexity of the aerial image and the large degree of freedom of remote sensing objects in scale, orientation, and density. In this paper, a novel multi-category rotation detector is proposed, which can efficiently detect small objects, arbitrary direction objects, and dense objects in complex remote sensing images. Specifically, the proposed model adopts a targeted feature fusion strategy called inception fusion network, which fully considers factors such as feature fusion, anchor sampling, and receptive field to improve the ability to handle small objects. Then we combine the pixel attention network and the channel attention network to weaken the noise information and highlight the objects feature. Finally, the rotational object detection algorithm is realized by redefining the rotating bounding box. Experiments on public datasets including DOTA, NWPU VHR-10 demonstrate that the proposed algorithm significantly outperforms state-of-the-art methods. The code and models will be available at this https URL.

84.Integrating domain knowledge: using hierarchies to improve deep classifiers pdf

One of the most prominent problems in machine learning in the age of deep learning is the availability of sufficiently large annotated datasets. While for standard problem domains (ImageNet classification), appropriate datasets exist, for specific domains, \eg classification of animal species, a long-tail distribution means that some classes are observed and annotated insufficiently. Challenges like iNaturalist show that there is a strong interest in species recognition. Acquiring additional labels can be prohibitively expensive. First, since domain experts need to be involved, and second, because acquisition of new data might be costly. Although there exist methods for data augmentation, which not always lead to better performance of the classifier, there is more additional information available that is to the best of our knowledge not exploited accordingly.
In this paper, we propose to make use of existing class hierarchies like WordNet to integrate additional domain knowledge into classification. We encode the properties of such a class hierarchy into a probabilistic model. From there, we derive a special label encoding together with a corresponding loss function. Using a convolutional neural network, on the ImageNet and NABirds datasets our method offers a relative improvement of 10.4% and 9.6% in accuracy over the baseline respectively. After less than a third of training time, it is already able to match the baseline's fine-grained recognition performance. Both results show that our suggested method is efficient and effective.

85.VommaNet: an End-to-End Network for Disparity Estimation from Reflective and Texture-less Light Field Images pdf

The precise combination of image sensor and micro-lens array enables lenslet light field cameras to record both angular and spatial information of incoming light, therefore, one can calculate disparity and depth from light field images. In turn, 3D models of the recorded objects can be recovered, which is a great advantage over other imaging system. However, reflective and texture-less areas in light field images have complicated conditions, making it hard to correctly calculate disparity with existing algorithms. To tackle this problem, we introduce a novel end-to-end network VommaNet to retrieve multi-scale features from reflective and texture-less regions for accurate disparity estimation. Meanwhile, our network has achieved similar or better performance in other regions for both synthetic light field images and real-world data compared to the state-of-the-art algorithms. Currently, we achieve the best score for mean squared error (MSE) on HCI 4D Light Field Benchmark.

86.Explicit Pose Deformation Learning for Tracking Human Poses pdf

We present a method for human pose tracking that learns explicitly about the dynamic effects of human motion on joint appearance. In contrast to previous techniques which employ generic tools such as dense optical flow or spatio-temporal smoothness constraints to pass pose inference cues between frames, our system instead learns to predict joint displacements from the previous frame to the current frame based on the possibly changing appearance of relevant pixels surrounding the corresponding joints in the previous frame. This explicit learning of pose deformations is formulated by incorporating concepts from human pose estimation into an optical flow-like framework. With this approach, state-of-the-art performance is achieved on standard benchmarks for various pose tracking tasks including 3D body pose tracking in RGB video, 3D hand pose tracking in depth sequences, and 3D hand gesture tracking in RGB video.

87.Not just a matter of semantics: the relationship between visual similarity and semantic similarity pdf

Knowledge transfer, zero-shot learning and semantic image retrieval are methods that aim at improving accuracy by utilizing semantic information, e.g. from WordNet. It is assumed that this information can augment or replace missing visual data in the form of labeled training images because semantic similarity somewhat aligns with visual similarity. This assumption may seem trivial, but is crucial for the application of such semantic methods. Any violation can cause mispredictions. Thus, it is important to examine the visual-semantic relationship for a certain target problem. In this paper, we use five different semantic and visual similarity measures each to thoroughly analyze the relationship without relying too much on any single definition. We postulate and verify three highly consequential hypotheses on the relationship. Our results show that it indeed exists and that WordNet semantic similarity carries more information about visual similarity than just the knowledge of "different classes look different". They suggest that classification is not the ideal application for semantic methods and that wrong semantic information is much worse than none.

88.Simulating LIDAR Point Cloud for Autonomous Driving using Real-world Scenes and Traffic Flows pdf

We present a LIDAR simulation framework that can automatically generate 3D point cloud based on LIDAR type and placement. The point cloud, annotated with ground truth semantic labels, is to be used as training data to improve environmental perception capabilities for autonomous driving vehicles. Different from previous simulators, we generate the point cloud based on real environment and real traffic flow. More specifically we employ a mobile LIDAR scanner with cameras to capture real world scenes. The input to our simulation framework includes dense 3D point cloud and registered color images. Moving objects (such as cars, pedestrians, bicyclists) are automatically identified and recorded. These objects are then removed from the input point cloud to restore a static background (e.g., environment without movable objects). With that we can insert synthetic models of various obstacles, such as vehicles and pedestrians in the static background to create various traffic scenes. A novel LIDAR renderer takes the composite scene to generate new realistic LIDAR points that are already annotated at point level for synthetic objects. Experimental results show that our system is able to close the performance gap between simulation and real data to be 1 ~ 6% in different applications, and for model fine tuning, only 10% ~ 20% extra real data could help to outperform the original model trained with full real dataset.

89.On Hallucinating Context and Background Pixels from a Face Mask using Multi-scale GANs pdf

We propose a multi-scale GAN model to hallucinate realistic context (forehead, hair, neck, clothes) and background pixels automatically from a single input face mask. Instead of swapping a face on to an existing picture, our model directly generates realistic context and background pixels based on the features of the provided face mask. Unlike face inpainting algorithms, it can generate realistic hallucinations even for a large number of missing pixels. Our model is composed of a cascaded network of GAN blocks, each tasked with hallucination of missing pixels at a particular resolution while guiding the synthesis process of the next GAN block. The hallucinated full face image is made photo-realistic by using a combination of reconstruction, perceptual, adversarial and identity preserving losses at each block of the network. With a set of extensive experiments, we demonstrate the effectiveness of our model in hallucinating context and background pixels from face masks varying in facial pose, expression and lighting, collected from multiple datasets subject disjoint with our training data. We also compare our method with two popular face swapping and face completion methods in terms of visual quality and recognition performance. Additionally, we analyze our cascaded pipeline and compare it with the recently proposed progressive growing of GANs.

90.Cross-modality deep learning brings bright-field microscopy contrast to holography pdf

Deep learning brings bright-field microscopy contrast to holographic images of a sample volume, bridging the volumetric imaging capability of holography with the speckle- and artifact-free image contrast of bright-field incoherent microscopy.

91.Deep Comparison: Relation Columns for Few-Shot Learning pdf

Few-shot deep learning is a topical challenge area for scaling visual recognition to open-ended growth in the space of categories to recognise. A promising line work towards realising this vision is deep networks that learn to match queries with stored training images. However, methods in this paradigm usually train a deep embedding followed by a single linear classifier. Our insight is that effective general-purpose matching requires discrimination with regards to features at multiple abstraction levels. We therefore propose a new framework termed \modelnamefull that decomposes embedding learning into a sequence of modules, and pairs each with a relation module. The relation modules compute a non-linear metric to score the match using the corresponding embedding module's representation. To ensure that all embedding module's features are used, the relation modules are deeply supervised. Finally generalisation is further improved by a learned noise regulariser. The resulting network achieves state of the art performance on both miniImageNet and tieredImageNet, while retaining the appealing simplicity and efficiency of deep metric learning approaches.

92.Alternating Segmentation and Simulation for Contrast Adaptive Tissue Classification pdf

A key feature of magnetic resonance (MR) imaging is its ability to manipulate how the intrinsic tissue parameters of the anatomy ultimately contribute to the contrast properties of the final, acquired image. This flexibility, however, can lead to substantial challenges for segmentation algorithms, particularly supervised methods. These methods require atlases or training data, which are composed of MR image and labeled image pairs. In most cases, the training data are obtained with a fixed acquisition protocol, leading to suboptimal performance when an input data set that requires segmentation has differing contrast properties. This drawback is increasingly significant with the recent movement towards multi-center research studies involving multiple scanners and acquisition protocols. In this work, we propose a new framework for supervised segmentation approaches that is robust to contrast differences between the training MR image and the input image. Our approach uses a generative simulation model within the segmentation process to compensate for the contrast differences. We allow the contrast of the MR image in the training data to vary by simulating a new contrast from the corresponding label image. The model parameters are optimized by a cost function measuring the consistency between the input MR image and its simulation based on a current estimate of the segmentation labels. We provide a proof of concept of this approach by combining a supervised classifier with a simple simulation model, and apply the resulting algorithm to synthetic images and actual MR images.

93.PydMobileNet: Improved Version of MobileNets with Pyramid Depthwise Separable Convolution pdf

Convolutional neural networks (CNNs) have shown remarkable performance in various computer vision tasks in recent years. However, the increasing model size has raised challenges in adopting them in real-time applications as well as mobile and embedded vision applications. Many works try to build networks as small as possible while still have acceptable performance. The state-of-the-art architecture is MobileNets. They use Depthwise Separable Convolution (DWConvolution) in place of standard Convolution to reduce the size of networks. This paper describes an improved version of MobileNet, called Pyramid Mobile Network. Instead of using just a $3\times 3$ kernel size for DWConvolution like in MobileNet, the proposed network uses a pyramid kernel size to capture more spatial information. The proposed architecture is evaluated on two highly competitive object recognition benchmark datasets (CIFAR-10, CIFAR-100). The experiments demonstrate that the proposed network achieves better performance compared with MobileNet as well as other state-of-the-art networks. Additionally, it is more flexible in fine-tuning the trade-off between accuracy, latency and model size than MobileNets.

94.Skeleton-based Gesture Recognition Using Several Fully Connected Layers with Path Signature Features and Temporal Transformer Module pdf

The skeleton based gesture recognition is gaining more popularity due to its wide possible applications. The key issues are how to extract discriminative features and how to design the classification model. In this paper, we first leverage a robust feature descriptor, path signature (PS), and propose three PS features to explicitly represent the spatial and temporal motion characteristics, i.e., spatial PS (S_PS), temporal PS (T_PS) and temporal spatial PS (T_S_PS). Considering the significance of fine hand movements in the gesture, we propose an "attention on hand" (AOH) principle to define joint pairs for the S_PS and select single joint for the T_PS. In addition, the dyadic method is employed to extract the T_PS and T_S_PS features that encode global and local temporal dynamics in the motion. Secondly, without the recurrent strategy, the classification model still faces challenges on temporal variation among different sequences. We propose a new temporal transformer module (TTM) that can match the sequence key frames by learning the temporal shifting parameter for each input. This is a learning-based module that can be included into standard neural network architecture. Finally, we design a multi-stream fully connected layer based network to treat spatial and temporal features separately and fused them together for the final result. We have tested our method on three benchmark gesture datasets, i.e., ChaLearn 2016, ChaLearn 2013 and MSRC-12. Experimental results demonstrate that we achieve the state-of-the-art performance on skeleton-based gesture recognition with high computational efficiency.

95.Weakly Supervised Semantic Image Segmentation with Self-correcting Networks pdf

Building a large image dataset with high-quality object masks for semantic segmentation is costly and time consuming. In this paper, we reduce the data preparation cost by leveraging weak supervision in the form of object bounding boxes. To accomplish this, we propose a principled framework that trains a deep convolutional segmentation model that combines a large set of weakly supervised images (having only object bounding box labels) with a small set of fully supervised images (having semantic segmentation labels and box labels). Our framework trains the primary segmentation model with the aid of an ancillary model that generates initial segmentation labels for the weakly supervised instances and a self-correction module that improves the generated labels during training using the increasingly accurate primary model. We introduce two variants of the self-correction module using either linear or convolutional functions. Experiments on the PASCAL VOC 2012 and Cityscape datasets show that our models trained with a small fully supervised set perform similar to, or better than, models trained with a large fully supervised set while requiring ~7x less annotation effort.

96.DSCnet: Replicating Lidar Point Clouds with Deep Sensor Cloning pdf

Convolutional neural networks (CNNs) have become increasingly popular for solving a variety of computer vision tasks, ranging from image classification to image segmentation. Recently, autonomous vehicles have created a demand for depth information, which is often obtained using hardware sensors such as Light detection and ranging (LIDAR). Although it can provide precise distance measurements, most LIDARs are still far too expensive to sell in mass-produced consumer vehicles, which has motivated methods to generate depth information from commodity automotive sensors like cameras.
In this paper, we propose an approach called Deep Sensor Cloning (DSC). The idea is to use Convolutional Neural Networks in conjunction with inexpensive sensors to replicate the 3D point-clouds that are created by expensive LIDARs. To accomplish this, we develop a new dataset (DSDepth) and a new family of CNN architectures (DSCnets). While previous tasks such as KITTI depth prediction use an interpolated RGB-D images as ground-truth for training, we instead use DSCnets to directly predict LIDAR point-clouds. When we compare the output of our models to a $75,000 LIDAR, we find that our most accurate DSCnet achieves a relative error of 5.77% using a single camera and 4.69% using stereo cameras.

97.Relational Long Short-Term Memory for Video Action Recognition pdf

Spatial and temporal relationships, both short-range and long-range, between objects in videos are key cues for recognizing actions. It is a challenging problem to model them jointly. In this paper, we first present a new variant of Long Short-Term Memory, namely Relational LSTM to address the challenge for relation reasoning across space and time between objects. In our Relational LSTM module, we utilize a non-local operation similar in spirit to the recently proposed non-local network to substitute the fully connected operation in the vanilla LSTM. By doing this, our Relational LSTM is capable of capturing long and short-range spatio-temporal relations between objects in videos in a principled way. Then, we propose a two-branch neural architecture consisting of the Relational LSTM module as the non-local branch and a spatio-temporal pooling based local branch. The local branch is introduced for capturing local spatial appearance and/or short-term motion features. The two-branch modules are concatenated to learn video-level features from snippet-level ones end-to-end. Experimental results on UCF-101 and HMDB-51 datasets show that our model achieves state-of-the-art results among LSTM-based methods, while obtaining comparable performance with other state-of-the-art methods (which use not directly comparable schema). Our code will be released.

98.Domain Adaptive Transfer Learning with Specialist Models pdf

Transfer learning is a widely used method to build high performing computer vision models. In this paper, we study the efficacy of transfer learning by examining how the choice of data impacts performance. We find that more pre-training data does not always help, and transfer performance depends on a judicious choice of pre-training data. These findings are important given the continued increase in dataset sizes. We further propose domain adaptive transfer learning, a simple and effective pre-training method using importance weights computed based on the target dataset. Our method to compute importance weights follow from ideas in domain adaptation, and we show a novel application to transfer learning. Our methods achieve state-of-the-art results on multiple fine-grained classification datasets and are well-suited for use in practice.

99.Improving Rotated Text Detection with Rotation Region Proposal Networks pdf

A significant number of images shared on social media platforms such as Facebook and Instagram contain text in various forms. It's increasingly becoming commonplace for bad actors to share misinformation, hate speech or other kinds of harmful content as text overlaid on images on such platforms. A scene-text understanding system should hence be able to handle text in various orientations that the adversary might use. Moreover, such a system can be incorporated into screen readers used to aid the visually impaired. In this work, we extend the scene-text extraction system at Facebook, Rosetta, to efficiently handle text in various orientations. Specifically, we incorporate the Rotation Region Proposal Networks (RRPN) in our text extraction pipeline and offer practical suggestions for building and deploying a model for detecting and recognizing text in arbitrary orientations efficiently. Experimental results show a significant improvement on detecting rotated text.

100.Topology-Aware Non-Rigid Point Cloud Registration pdf

In this paper, we introduce a non-rigid registration pipeline for pairs of unorganized point clouds that may be topologically different. Standard warp field estimation algorithms, even under robust, discontinuity-preserving regularization, tend to produce erratic motion estimates on boundaries associated with close-to-open' topology changes. We overcome this limitation by exploiting backward motion: in the opposite motion direction, a close-to-open' event becomes open-to-close', which is by default handled correctly. At the core of our approach lies a general, topology-agnostic warp field estimation algorithm, similar to those employed in recently introduced dynamic reconstruction systems from RGB-D input. We improve motion estimation on boundaries associated with topology changes in an efficient post-processing phase. Based on both forward and (inverted) backward warp hypotheses, we explicitly detect regions of the deformed geometry that undergo topological changes by means of local deformation criteria and broadly classify them as contacts' or `separations'. Subsequently, the two motion hypotheses are seamlessly blended on a local basis, according to the type and proximity of detected events. Our method achieves state-of-the-art motion estimation accuracy on the MPI Sintel dataset. Experiments on a custom dataset with topological event annotations demonstrate the effectiveness of our pipeline in estimating motion on event boundaries, as well as promising performance in explicit topological event detection.

101.Coupling weak and strong supervision for classification of prostate cancer histopathology images pdf

Automated grading of prostate cancer histopathology images is a challenging task, with one key challenge being the scarcity of annotations down to the level of regions of interest (strong labels), as typically the prostate cancer Gleason score is known only for entire tissue slides (weak labels). In this study, we focus on automated Gleason score assignment of prostate cancer whole-slide images on the basis of a large weakly-labeled dataset and a smaller strongly-labeled one. We efficiently leverage information from both label sources by jointly training a classifier on the two datasets and by introducing a gradient update scheme that assigns different relative importances to each training example, as a means of self-controlling the weak supervision signal. Our approach achieves superior performance when compared with standard Gleason scoring methods.

102.Data-Efficient Graph Embedding Learning for PCB Component Detection pdf

This paper presents a challenging computer vision task, namely the detection of generic components on a PCB, and a novel set of deep-learning methods that are able to jointly leverage the appearance of individual components and the propagation of information across the structure of the board to accurately detect and identify various types of components on a PCB. Due to the expense of manual data labeling, a highly unbalanced distribution of component types, and significant domain shift across boards, most earlier attempts based on traditional image processing techniques fail to generalize well to PCB images with various quality, lighting conditions, etc. Newer object detection pipelines such as Faster R-CNN, on the other hand, require a large amount of labeled data, do not deal with domain shift, and do not leverage structure. To address these issues, we propose a three stage pipeline in which a class-agnostic region proposal network is followed by a low-shot similarity prediction classifier. In order to exploit the data dependency within a PCB, we design a novel Graph Network block to refine the component features conditioned on each PCB. To the best of our knowledge, this is one of the earliest attempts to train a deep learning based model for such tasks, and we demonstrate improvements over recent graph networks for this task. We also provide in-depth analysis and discussion for this challenging task, pointing to future research.

103.Adversarial Autoencoders for Generating 3D Point Clouds pdf

Deep generative architectures provide a way to model not only images, but also complex, 3-dimensional objects, such as point clouds. In this work, we present a novel method to obtain meaningful representations of 3D shapes that can be used for clustering and reconstruction. Contrary to existing methods for 3D point cloud generation that train separate decoupled models for representation learning and generation, our approach is the first end-to-end solution that allows to simultaneously learn a latent space of representation and generate 3D shape out of it. To achieve this goal, we extend a deep Adversarial Autoencoder model (AAE) to accept 3D input and create 3D output. Thanks to our end-to-end training regime, the resulting method called 3D Adversarial Autoencoder (3dAAE) obtains either binary or continuous latent space that covers much wider portion of training data distribution, hence allowing smooth interpolation between the shapes. Finally, our extensive quantitative evaluation shows that 3dAAE provides state-of-the-art results on a set of benchmark tasks.

104.Three Dimensional Convolutional Neural Network Pruning with Regularization-Based Method pdf

In recent years, three-dimensional convolutional neural network (3D CNN) is intensively applied in video analysis and receives good performance. However, 3D CNN leads to massive computation and storage consumption, which hinders its deployment on mobile and embedded devices. In this paper, we propose a three-dimensional regularization-based pruning method to assign different regularization parameters to different weight groups based on their importance to the network. Experiments show that the proposed method outperforms other popular methods in this area.

105.Bayesian CycleGAN via Marginalizing Latent Sampling pdf

Recent techniques built on Generative Adversarial Networks (GANs) like CycleGAN are able to learn mappings between domains from unpaired datasets through min-max optimization games between generators and discriminators. However, it remains challenging to stabilize training process and diversify generated results. To address these problems, we present a Bayesian extension of cyclic model and an integrated cyclic framework for inter-domain mappings. The proposed method stimulated by Bayesian GAN explores the full posteriors of Bayesian cyclic model (with latent sampling) and optimizes the model with maximum a posteriori (MAP) estimation. Hence, we name it {\tt Bayesian CycleGAN}. We perform the proposed Bayesian CycleGAN on multiple benchmark datasets, including Cityscapes, Maps, and Monet2photo. The quantitative and qualitative evaluations demonstrate the proposed method can achieve more stable training, superior performance and diversified images generating.

106.PerSIM: Multi-resolution Image Quality Assessment in the Perceptually Uniform Color Domain pdf

An average observer perceives the world in color instead of black and white. Moreover, the visual system focuses on structures and segments instead of individual pixels. Based on these observations, we propose a full reference objective image quality metric modeling visual system characteristics and chroma similarity in the perceptually uniform color domain (Lab). Laplacian of Gaussian features are obtained in the L channel to model the retinal ganglion cells in human visual system and color similarity is calculated over the a and b channels. In the proposed perceptual similarity index (PerSIM), a multi-resolution approach is followed to mimic the hierarchical nature of human visual system. LIVE and TID2013 databases are used in the validation and PerSIM outperforms all the compared metrics in the overall databases in terms of ranking, monotonic behavior and linearity.

107.Distribution Discrepancy Maximization for Image Privacy Preserving pdf

With the rapid increase in online photo sharing activities, image obfuscation algorithms become particularly important for protecting the sensitive information in the shared photos. However, existing image obfuscation methods based on hand-crafted principles are challenged by the dramatic development of deep learning techniques. To address this problem, we propose to maximize the distribution discrepancy between the original image domain and the encrypted image domain. Accordingly, we introduce a collaborative training scheme: a discriminator $D$ is trained to discriminate the reconstructed image from the encrypted image, and an encryption model $G_e$ is required to generate these two kinds of images to maximize the recognition rate of $D$, leading to the same training objective for both $D$ and $G_e$. We theoretically prove that such a training scheme maximizes two distributions' discrepancy. Compared with commonly-used image obfuscation methods, our model can produce satisfactory defense against the attack of deep recognition models indicated by significant accuracy decreases on FaceScrub, Casia-WebFace and LFW datasets.

108.Regularized adversarial examples for model interpretability pdf

As machine learning algorithms continue to improve, there is an increasing need for explaining why a model produces a certain prediction for a certain input. In recent years, several methods for model interpretability have been developed, aiming to provide explanation of which subset regions of the model input is the main reason for the model prediction. In parallel, a significant research community effort is occurring in recent years for developing adversarial example generation methods for fooling models, while not altering the true label of the input,as it would have been classified by a human annotator. In this paper, we bridge the gap between adversarial example generation and model interpretability, and introduce a modification to the adversarial example generation process which encourages better interpretability. We analyze the proposed method on a public medical imaging dataset, both quantitatively and qualitatively, and show that it significantly outperforms the leading known alternative method. Our suggested method is simple to implement, and can be easily plugged into most common adversarial example generation frameworks. Additionally, we propose an explanation quality metric - $APE$ - "Adversarial Perturbative Explanation", which measures how well an explanation describes model decisions.

109.Enhancing the Robustness of Prior Network in Out-of-Distribution Detection pdf

With the recent surge of interests in deep neural networks, more real-world applications start to adopt it in practice. However, deep neural networks are known to have limited control over its prediction under unseen images. Such weakness can potentially threaten society and cause annoying consequences in real-world scenarios. In order to resolve such issue, a popular task called out-of-distribution detection was proposed, which aims at separating out-of-distribution images from in-distribution images. In this paper, we propose a perturbed prior network architecture, which can efficiently separate model-level uncertainty from data-level uncertainty via prior entropy. To further enhance the robustness of proposed entropy-based uncertainty measure, we propose a concentration perturbation algorithm, which adaptively adds noise to concentration parameters so that the in- and out-of-distribution images are better separable. Our method can directly rely on the pre-trained deep neural network without re-training it, and also requires no knowledge about the network architecture and out-of-distribution examples. Such simplicity makes our method more suitable for real-world AI applications. Through comprehensive experiments, our methods demonstrate its superiority by achieving state-of-the-art results on many datasets.

110.GAN-QP: A Novel GAN Framework without Gradient Vanishing and Lipschitz Constraint pdf

We know SGAN may have a risk of gradient vanishing. A significant improvement is WGAN, with the help of 1-Lipschitz constraint on discriminator to prevent from gradient vanishing. Is there any GAN having no gradient vanishing and no 1-Lipschitz constraint on discriminator? We do find one, called GAN-QP.
To construct a new framework of Generative Adversarial Network (GAN) usually includes three steps: 1. choose a probability divergence; 2. convert it into a dual form; 3. play a min-max game. In this articles, we demonstrate that the first step is not necessary. We can analyse the property of divergence and even construct new divergence in dual space directly. As a reward, we obtain a simpler alternative of WGAN: GAN-QP. We demonstrate that GAN-QP have a better performance than WGAN in theory and practice.

111.Classifiers Based on Deep Sparse Coding Architectures are Robust to Deep Learning Transferable Examples pdf

Although deep learning has shown great success in recent years, researchers have discovered a critical flaw where small, imperceptible changes in the input to the system can drastically change the output classification. These attacks are exploitable in nearly all of the existing deep learning classification frameworks. However, the susceptibility of deep sparse coding models to adversarial examples has not been examined. Here, we show that classifiers based on a deep sparse coding model whose classification accuracy is competitive with a variety of deep neural network models are robust to adversarial examples that effectively fool those same deep learning models. We demonstrate both quantitatively and qualitatively that the robustness of deep sparse coding models to adversarial examples arises from two key properties. First, because deep sparse coding models learn general features corresponding to generators of the dataset as a whole, rather than highly discriminative features for distinguishing specific classes, the resulting classifiers are less dependent on idiosyncratic features than might be more easily exploited. Second, because deep sparse coding models utilize fixed point attractor dynamics with top-down feedback, it is more difficult to find small changes to the input that drive the resulting representations out of the correct attractor basin.

112.BLeSS: Bio-inspired Low-level Spatiochromatic Similarity Assisted Image Quality Assessment pdf

This paper proposes a biologically-inspired low-level spatiochromatic-model-based similarity method (BLeSS) to assist full-reference image-quality estimators that originally oversimplify color perception processes. More specifically, the spatiochromatic model is based on spatial frequency, spatial orientation, and surround contrast effects. The assistant similarity method is used to complement image-quality estimators based on phase congruency, gradient magnitude, and spectral residual. The effectiveness of BLeSS is validated using FSIM, FSIMc and SR-SIM methods on LIVE, Multiply Distorted LIVE, and TID 2013 databases. In terms of Spearman correlation, BLeSS enhances the performance of all quality estimators in color-based degradations and the enhancement is at 100% for both feature- and spectral residual-based similarity methods. Moreover, BleSS significantly enhances the performance of SR-SIM and FSIM in the full TID 2013 database.

113.An Infinite Parade of Giraffes: Expressive Augmentation and Complexity Layers for Cartoon Drawing pdf

In this paper, we explore creative image generation constrained by small data. To partially automate the creation of cartoon sketches consistent with a specific designer's style, where acquiring a very large original image set is impossible or cost prohibitive, we exploit domain specific knowledge for a huge reduction in original image requirements, creating an effectively infinite number of cartoon giraffes from just nine original drawings. We introduce "expressive augmentations" for cartoon sketches, mathematical transformations that create broad domain appropriate variation, far beyond the usual affine transformations, and we show that chained GANs models trained on the temporal stages of drawing or "complexity layers" can effectively add character appropriate details and finish new drawings in the designer's style.
We discuss the application of these tools in design processes for textiles, graphics, architectural elements and interior design.

114.Learned Video Compression pdf

We present a new algorithm for video coding, learned end-to-end for the low-latency mode. In this setting, our approach outperforms all existing video codecs across nearly the entire bitrate range. To our knowledge, this is the first ML-based method to do so.
We evaluate our approach on standard video compression test sets of varying resolutions, and benchmark against all mainstream commercial codecs, in the low-latency mode. On standard-definition videos, relative to our algorithm, HEVC/H.265, AVC/H.264 and VP9 typically produce codes up to 60% larger. On high-definition 1080p videos, H.265 and VP9 typically produce codes up to 20% larger, and H.264 up to 35% larger. Furthermore, our approach does not suffer from blocking artifacts and pixelation, and thus produces videos that are more visually pleasing.
We propose two main contributions. The first is a novel architecture for video compression, which (1) generalizes motion estimation to perform any learned compensation beyond simple translations, (2) rather than strictly relying on previously transmitted reference frames, maintains a state of arbitrary information learned by the model, and (3) enables jointly compressing all transmitted signals (such as optical flow and residual).
Secondly, we present a framework for ML-based spatial rate control: namely, a mechanism for assigning variable bitrates across space for each frame. This is a critical component for video coding, which to our knowledge had not been developed within a machine learning setting.