Pytorch for Classification and Retrieval
Please refer to get_started.md for installation.
TODO
- openCV
- PIL
- Imbalance
- general Augmentaions and albumentations
- CutOut/Mixup/CutMix/SnapMix
(fork from TIMM)
- TIMM
- dropblock
- Non local
- DCN
- SKNet
- CBAM
- RGA Module
- GAP
- GMP
- GAP + GMP
- GeM(w/o Fix)
- SoftPool
- BNNeck Head
- Identity Head
- Reduction Head
- CE
- FocalLoss
- Norm Softmax
- Triplet
- CosFace
- ArcFace
- SnapMix
- EqLoss
- Circle
- TensorBoard
- EMA
- Multi-Scale Training
- w/o Multi-GPUs
- TTA
- Normalize
- PCA
- DBA
- QE(w/o weight)
- Euclidean
- Cosine
- K-reciprocal
- K-reciprocal++
- SA Modules
- Noisy Label
- More Data Augmentaions
- More Convenient
[1] FastReID
[2] PyRetri
[3] mmdetection
