Skip to content
Real-time 'Actor-Critic' tracking
Branch: master
Clone or download
Latest commit 8d86e86 Jan 1, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
dataset/Car4 test code Sep 18, 2018
models test code Sep 18, 2018
modules test code Sep 18, 2018
tracking remove directory from remote repository Sep 18, 2018
train Training code Sep 20, 2018
ACT.png
LICENSE Create LICENSE Oct 18, 2018
readme.md Update readme.md Nov 5, 2018

readme.md

Real-time 'Actor-Critic' tracking

Code for Real-time 'Actor-Critic' tracking accepted by ECCV 2018

Introduction

We propose a novel tracking algorithm with real-time performance based on the ‘Actor-Critic’ framework.
Results on OTB100

Requirements

  1. Tensorflow 1.4.0 (Train) and Pytorch 0.3.0 (Test)
  2. CUDA 8.0 and cuDNN 6.0
  3. Python 2.7

Usage

Train

  1. Please download the ILSVRC VID dataset, and put the VID folder into $(ACT_root)/train/
    (We adopt the same videos as meta_trackers. You can find more details in ilsvrc_train.json.)
  2. Run the $(ACT_root)/train/DDPG_train.py to train the 'Actor and Critic' network.

Test

Please run $(ACT_root)/tracking/run_tracker.py for demo.

License

Licensed under an MIT license.

Citation

If you find ACT useful in your research, please kindly cite our paper:

@InProceedings{Chen_2018_ECCV,
author = {Chen, Boyu and Wang, Dong and Li, Peixia and Wang, Shuang and Lu, Huchuan},
title = {Real-time 'Actor-Critic' Tracking},
booktitle = {The European Conference on Computer Vision (ECCV)},
month = {September},
year = {2018}
}

Contact

If you have any questions, please feel free to contact bychen@mail.dlut.edu.cn

Acknowledgments

Many parts of this code are adopted from other related works (py-MDNet and meta_trackers)

You can’t perform that action at this time.