This work is published in Pattern Recognition. Please cite the following paper should you consider to use this code.
- Xiaohan Yu, Yang Zhao, Yongsheng Gao, Shengwu Xiong (2021). MaskCOV: A Random Mask Covariance Network for Ultra-Fine-Grained Visual Categorization. In Pattern Recognition.
@article{yu2021maskcov, title={MaskCOV: A Random Mask Covariance Network for Ultra-Fine-Grained Visual Categorization}, author={Yu, Xiaohan and Zhao, Yang and Gao, Yongsheng and Xiong, Shengwu}, journal={Pattern Recognition}, pages={108067}, year={2021}, publisher={Elsevier} }
Please find our code in the folder PR_MaskCOV. The ultra-fine-grained image dataset, UFG, used in this paper can be downloaded via "https://github.com/XiaohanYu-GU/Ultra-FGVC".
install pytorch 1.6.0, python 3.7, cuda 10.1, cudnn7.6.3 and any necessary python package that is required.
Use the following order to run the training code in a default setting.
"sh main.sh"
Or revise the hyper-parameters (batch size, learning rate) in config.py if needed and then run "sh main.sh".
For Cotton80 subset, the batch size is recommended to be 8. For the remaining subsets, the batch size is recommended to be 16.
The code is revised based on source code provided by DCL (see "https://github.com/JDAI-CV/DCL"). We sincerely thank their contribution.
Xiaohan Yu, yuxiaohan112@gmail.com