Code for paper "Hetero-center loss for cross-modality person re-identification"
2020-06-03： Because pytorch split the data of a batch to each gpu when using multi-gpus, the loss may be incorrectly computed. The users may need to use single gpu to reproduce the experimental results in the paper.
2020-01-07: we fix up a bug in learning rate schedule, before that only the first three parameter group's learning rate will be correctly decay to 1/10. However, after fixing up the bug, the model's performance still stay the same. The updated model and code have been upload.
pytorch: 0.4.1(the higher version may lead to performance fluctuation)
torchvision: 0.2.1
numpy: 1.17.4
python: 3.7
SYSU-MM01
Reg-DB
- prepare training set
python pre_process_sysu.py
- train model
python train.py --dataset sysu --lr 0.01 --drop 0.0 --trial 1 --gpu 1 --epochs 60 --w_hc 0.5 --per_img 8
- (Notice that you need to set the 88 line in train.py to your SYSU-MM01 dataset path)
- evaluate model(single-shot all-search)
python test.py --dataset sysu --lr 0.01 --drop 0.0 --trial 1 --gpu 1 --low-dim 512 --resume 'Your model name' --w_hc 0.5 --mode all --gall-mode single --model_path 'Your model path'
- train model
python train.py --dataset regdb --lr 0.01 --drop 0.0 --trial 1 --gpu 1 --epochs 60 --w_hc 0.5 --per_img 8
- evaluate model
python test.py --dataset regdb --lr 0.01 --drop 0.0 --trial 1 --gpu 1 --low-dim 512 --resume 'Your model name' --w_hc 0.5 --model_path 'Your model path'
| Dataset | Rank1 | mAP | model |
|---|---|---|---|
| SYSU-MM01 | ~56% | ~54% | BaiduYun(code:y2em) |
| RegDB | ~83% | ~72% | BaiduYun(code:y2em) |
Because this is the first time I use Github to release my code, maybe this project is a little difficult to read and use. If you have any question, please don't hesitate to contact me (zhuyuanxin98@outlook.com). I will reply to you as soon as possible.
Most of the code are borrowed from https://github.com/mangye16/Cross-Modal-Re-ID-baseline. I am very grateful to the author (@mangye16) for his contribution and help.
If you think this project useful, please give me a star and cite following papers:
[1] Zhu Y, Yang Z, Wang L, et al. Hetero-Center Loss for Cross-Modality Person Re-Identification[J]. Neurocomputing, 2019.
[2] Ye M, Lan X, Wang Z, et al. Bi-directional Center-Constrained Top-Ranking for Visible Thermal Person Re-Identification[J]. IEEE Transactions on Information Forensics and Security, 2019.