Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/STVIR/pysot
SenseTime Research platform for single object tracking, implementing algorithms like SiamRPN and SiamMask.
https://github.com/STVIR/pysot
computer-vision siamese-network tracking
Last synced: about 2 months ago
JSON representation
SenseTime Research platform for single object tracking, implementing algorithms like SiamRPN and SiamMask.
- Host: GitHub
- URL: https://github.com/STVIR/pysot
- Owner: STVIR
- License: apache-2.0
- Created: 2019-05-07T12:37:56.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2023-11-12T03:26:20.000Z (10 months ago)
- Last Synced: 2024-05-09T10:32:59.637Z (4 months ago)
- Topics: computer-vision, siamese-network, tracking
- Language: Python
- Homepage:
- Size: 6.2 MB
- Stars: 4,380
- Watchers: 163
- Forks: 1,112
- Open Issues: 55
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome-robotic-tooling - pysot - The goal of PySOT is to provide a high-quality, high-performance codebase for visual tracking research. (Sensor Processing / Image Processing)
README
# PySOT
**PySOT** is a software system designed by SenseTime Video Intelligence Research team. It implements state-of-the-art single object tracking algorithms, including [SiamRPN](http://openaccess.thecvf.com/content_cvpr_2018/html/Li_High_Performance_Visual_CVPR_2018_paper.html) and [SiamMask](https://arxiv.org/abs/1812.05050). It is written in Python and powered by the [PyTorch](https://pytorch.org) deep learning framework. This project also contains a Python port of toolkit for evaluating trackers.
PySOT has enabled research projects, including: [SiamRPN](http://openaccess.thecvf.com/content_cvpr_2018/html/Li_High_Performance_Visual_CVPR_2018_paper.html), [DaSiamRPN](https://arxiv.org/abs/1808.06048), [SiamRPN++](https://arxiv.org/abs/1812.11703), and [SiamMask](https://arxiv.org/abs/1812.05050).
Example SiamFC, SiamRPN and SiamMask outputs.
## Introduction
The goal of PySOT is to provide a high-quality, high-performance codebase for visual tracking *research*. It is designed to be flexible in order to support rapid implementation and evaluation of novel research. PySOT includes implementations of the following visual tracking algorithms:
- [SiamMask](https://arxiv.org/abs/1812.05050)
- [SiamRPN++](https://arxiv.org/abs/1812.11703)
- [DaSiamRPN](https://arxiv.org/abs/1808.06048)
- [SiamRPN](http://openaccess.thecvf.com/content_cvpr_2018/html/Li_High_Performance_Visual_CVPR_2018_paper.html)
- [SiamFC](https://arxiv.org/abs/1606.09549)
using the following backbone network architectures:
- [ResNet{18, 34, 50}](https://arxiv.org/abs/1512.03385)
- [MobileNetV2](https://arxiv.org/abs/1801.04381)
- [AlexNet](https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks)
Additional backbone architectures may be easily implemented. For more details about these models, please see [References](#references) below.
Evaluation toolkit can support the following datasets:
:paperclip: [OTB2015](http://faculty.ucmerced.edu/mhyang/papers/pami15_tracking_benchmark.pdf)
:paperclip: [VOT16/18/19](http://votchallenge.net)
:paperclip: [VOT18-LT](http://votchallenge.net/vot2018/index.html)
:paperclip: [LaSOT](https://arxiv.org/pdf/1809.07845.pdf)
:paperclip: [UAV123](https://arxiv.org/pdf/1804.00518.pdf)
## Model Zoo and Baselines
We provide a large set of baseline results and trained models available for download in the [PySOT Model Zoo](MODEL_ZOO.md).
## Installation
Please find installation instructions for PyTorch and PySOT in [`INSTALL.md`](INSTALL.md).
## Quick Start: Using PySOT
### Add PySOT to your PYTHONPATH
```bash
export PYTHONPATH=/path/to/pysot:$PYTHONPATH
```
### Download models
Download models in [PySOT Model Zoo](MODEL_ZOO.md) and put the model.pth in the correct directory in experiments
### Webcam demo
```bash
python tools/demo.py \
--config experiments/siamrpn_r50_l234_dwxcorr/config.yaml \
--snapshot experiments/siamrpn_r50_l234_dwxcorr/model.pth
# --video demo/bag.avi # (in case you don't have webcam)
```
### Download testing datasets
Download datasets and put them into `testing_dataset` directory. Jsons of commonly used datasets can be downloaded from [Google Drive](https://drive.google.com/drive/folders/10cfXjwQQBQeu48XMf2xc_W1LucpistPI) or [BaiduYun](https://pan.baidu.com/s/1js0Qhykqqur7_lNRtle1tA#list/path=%2F). If you want to test tracker on new dataset, please refer to [pysot-toolkit](https://github.com/StrangerZhang/pysot-toolkit) to setting `testing_dataset`.
### Test tracker
```bash
cd experiments/siamrpn_r50_l234_dwxcorr
python -u ../../tools/test.py \
--snapshot model.pth \ # model path
--dataset VOT2018 \ # dataset name
--config config.yaml # config file
```
The testing results will in the current directory(results/dataset/model_name/)
### Eval tracker
assume still in experiments/siamrpn_r50_l234_dwxcorr_8gpu
``` bash
python ../../tools/eval.py \
--tracker_path ./results \ # result path
--dataset VOT2018 \ # dataset name
--num 1 \ # number thread to eval
--tracker_prefix 'model' # tracker_name
```
### Training :wrench:
See [TRAIN.md](TRAIN.md) for detailed instruction.
### Getting Help :hammer:
If you meet problem, try searching our GitHub issues first. We intend the issues page to be a forum in which the community collectively troubleshoots problems. But please do **not** post **duplicate** issues. If you have similar issue that has been closed, you can reopen it.
- `ModuleNotFoundError: No module named 'pysot'`
:dart:Solution: Run `export PYTHONPATH=path/to/pysot` first before you run the code.
- `ImportError: cannot import name region`
:dart:Solution: Build `region` by `python setup.py build_ext —-inplace` as decribled in [INSTALL.md](INSTALL.md).
## References
- [Fast Online Object Tracking and Segmentation: A Unifying Approach](https://arxiv.org/abs/1812.05050).
Qiang Wang, Li Zhang, Luca Bertinetto, Weiming Hu, Philip H.S. Torr.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019.
- [SiamRPN++: Evolution of Siamese Visual Tracking with Very Deep Networks](https://arxiv.org/abs/1812.11703).
Bo Li, Wei Wu, Qiang Wang, Fangyi Zhang, Junliang Xing, Junjie Yan.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019.
- [Distractor-aware Siamese Networks for Visual Object Tracking](https://arxiv.org/abs/1808.06048).
Zheng Zhu, Qiang Wang, Bo Li, Wu Wei, Junjie Yan, Weiming Hu.
The European Conference on Computer Vision (ECCV), 2018.
- [High Performance Visual Tracking with Siamese Region Proposal Network](http://openaccess.thecvf.com/content_cvpr_2018/html/Li_High_Performance_Visual_CVPR_2018_paper.html).
Bo Li, Wei Wu, Zheng Zhu, Junjie Yan, Xiaolin Hu.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018.
- [Fully-Convolutional Siamese Networks for Object Tracking](https://arxiv.org/abs/1606.09549).
Luca Bertinetto, Jack Valmadre, João F. Henriques, Andrea Vedaldi, Philip H. S. Torr.
The European Conference on Computer Vision (ECCV) Workshops, 2016.
## Contributors
- [Fangyi Zhang](https://github.com/StrangerZhang)
- [Qiang Wang](http://www.robots.ox.ac.uk/~qwang/)
- [Bo Li](http://bo-li.info/)
- [Zhiyuan Chen](https://zyc.ai/)
- [Jinghao Zhou](https://shallowtoil.github.io/)
## License
PySOT is released under the [Apache 2.0 license](https://github.com/STVIR/pysot/blob/master/LICENSE).