This repository provides a PyTorch implementation of the paper Structured Context Enhancement Network for Mouse Pose Estimation.
Tested with:
-
PyTorch 1.4.0
-
Torchvision 0.5.0
-
Python 3.6.8
- Download the data from DeepLabCut Mouse Pose and DeepPoseKit Animal Pose. Then put them under the data directory:
-labeled-data\ -mouse\ ... -flyimage\ ... -zebraimage\ ...
- Before running
train.py
, we need to compile Directionmax operation used in our paper, which is inspired by the corner pooling scheme in CornerNet.
`cd <CornerNet dir>/models/py_utils/_cpools/`
`python setup.py install --user`
- Then train the model
`python train.py`
All experimental procedures were performed in accordance with the Guidance on the Operation of the Animals (Scientific Procedures) Act, 1986 (UK) and approved by the Queen’s University Belfast Animal.
If you find this repository useful, please cite our paper:
@article{zhou2021structured,
title={Structured Context Enhancement Network for Mouse Pose Estimation},
author={Zhou, Feixiang and Jiang, Zheheng and Liu, Zhihua and Chen, Fang and Chen, Long and Tong, Lei and Yang, Zhile and Wang, Haikuan and Fei, Minrui and Li, Ling and others},
journal={IEEE Transactions on Circuits and Systems for Video Technology},
year={2021},
publisher={IEEE}
}
For any questions, feel free to contact: fz64@leicester.ac.uk