• Stars
    star
    133
  • Rank 272,600 (Top 6 %)
  • Language
    Python
  • Created over 5 years ago
  • Updated about 5 years ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

Learning the Model Update for Siamese Trackers (ICCV 2019)

Learning the Model Update for Siamese Trackers

The paper will appear in ICCV 2019.

arXiv, ICCV2019_openaccess, supplementary material, poster.

Citation

Please cite our paper if you are inspired by this idea.

@InProceedings{Zhang_2019_ICCV,
author = {Zhang, Lichao and Gonzalez-Garcia, Abel and Weijer, Joost van de and Danelljan, Martin and Khan, Fahad Shahbaz},
title = {Learning the Model Update for Siamese Trackers},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {October},
year = {2019}
} 

Instructions

In general, in Siamese trackers the template is linearly combined with the accumulated template from the previous frame, resulting in an exponential decay of information over time. While such an approach to updating has led to improved results, its simplicity limits the potential gain likely to be obtained by learning to update. Therefore, in this paper we propose to replace the handcrafted update function with a method which learns to update. We use a convolutional neural network, called UpdateNet, which given the initial template, the accumulated template and the template of the current frame aims to estimate the optimal template for the next frame. The UpdateNet is compact and can easily be integrated into existing Siamese trackers, e.g. SiamFC [1] and DaSiamRPN [2].

Visualization of learned accumulated templates and response maps.


Framework

The released project is done in PyTorch framework.

Pre-trained models

The pre-trained models are available to download.

Raw results

The results are available for comparison.

[1] Luca Bertinetto, Jack Valmadre, JoΓ£o F. Henriques, Andrea Vedaldi, Philip H. S. Torr. Fully-convolutional siamese networks for object tracking. In ECCV workshop 2016.

[2] Zheng Zhu, Qiang Wang, Bo Li, Wei Wu, Junjie Yan, Weiming Hu. Distractor-aware siamese networks for visual object tracking. In ECCV 2018.