ADNet-cvpr2017

• ADNet: Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning [Paper]
• [Project Page]
• Contact: yunsd101@snu.ac.kr [Homepage]

Other implementations

Tensorflow: (tf-adnet) https://github.com/ildoonet/tf-adnet-tracking

News

• Major bug fixes (31 Mar. 2018).
• Training codes are uploaded (4 Sep, 2017).
• Test codes are uploaded (15 July, 2017).

Citation

  @InProceedings{yun2017adnet,
  title={Action-Decision Networks for Visual Tracking with Deep Reinforcement Learning},
  author={Yun, Sangdoo and Choi, Jongwon and Yoo, Youngjoon and Yun, Kimin and Young Choi, Jin},
  booktitle={The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},  
  year={2017}
  }

Installation

• This code is tested on Linux os (ubuntu 14.04) 64bit, MATLAB 2017a, and Cuda-8.0 with NVIDIA GTX 1080 TI.
• Run adnet_compile.m to compile MatConvNet.

Run tracking

• Run adnet_demo.m.