- Classification: reid_classification.py
- Classification + triplet loss: reid_tripletcls.py (triplet_loss)
- Classification + triplet loss with hard negative mining: reid_tripletcls.py (triplet_hard_loss)
- Re-ranking with k-reciprocal Encoding (CVPR2017): re_ranking.py
- Pre-trained model: naivehard_more_last.h5 (market1501:81.1% rank-1 accuracy cuhk03:78.8% rank-1 accuracy)
http://pan.baidu.com/s/1bo3gwaV
You should add some functions according to the comments