Spatial Pyramid Attention for Deep Convolutional Neural Networks

Xu Ma, Jingda Guo, Andrew Sansom, Mara McGuire, Andrew Kalaani, Qi Chen, Sihai Tang, Qing Yang, Song Fu*

SPA module

Getting Start

Installation

1. Download repo

git clone https://github.com/13952522076/SPANet_TMM.git
cd SPANet_TMM

2. Requirements

• Python3.6
• PyTorch 1.3+
• CUDA 10+
• GCC 5.0+

pip install -r requirements.txt

3. Install DALI and Apex （For ImageNet Training）

DALI Installation:

cd ~
# For CUDA10
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist nvidia-dali-cuda100
# or
# For CUDA11
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist nvidia-dali-cuda110

For more details, please see Nvidia DALI installation.

Apex Installation:

cd ~
git clone https://github.com/NVIDIA/apex
cd apex
pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" ./

For more details, please see Apex or Apex Full API documentation.

Training & Testing ImageNet

We provide two training strategies: step_lr schedular and cosine_lr schedular in main.py and main_mobile.py respectively.

The training models (last one and best one) and the log file are saved in "checkpoints/imagenet/model_name" by default.

---

I personally suggest to manually setup the path to imagenet dataset in main.py (line 49) and main_mobile.py (line 50). Replace the default value to your real PATH.

Or you can add a parameter --data in the following training command.

For the step learning rate schedular, run follwing commands

# change the parameters accordingly if necessary
# e.g, If you have 4 GPUs, set the nproc_per_node to 4. If you want to train with 32FP, remove ----fp16.
python3 -m torch.distributed.launch --nproc_per_node=8 main.py -a spa_resnet50 --fp16 --b 32

For the cosine learning rate schedular, run follwing commands

# change the parameters accordingly if necessary
python3 -m torch.distributed.launch --nproc_per_node=8 main_mobile.py -a spa_resnet18 --b 64 --opt-level O0

Training & Testing Downsampled ImageNet

python3 downsample.py --netName=SPAResNet18 --bs=512

Training & Testing CIFAR

python3 cifar.py --netName=SPAResNet18 --cifar=100 --bs=512

Calculate Parameters and FLOPs

python3 count_Param.py

Citation

@inproceedings{SPA2020Xu,
  title={Spanet: Spatial Pyramid Attention Network for Enhanced Image Recognition},
  author={Guo, Jingda and Ma, Xu and Sansom, Andrew and McGuire, Mara and Kalaani, Andrew and Chen, Qi and Tang, Sihai and Yang, Qing and Fu, Song},
  booktitle={2020 IEEE International Conference on Multimedia and Expo (ICME)},
  pages={1--6},
  year={2020},
  organization={IEEE}
}