mmpretrain/projects/fgia_accv2022_1st

Solution of FGIA ACCV 2022(1st Place)

This is fine-tuning part of the 1st Place Solution for Webly-supervised Fine-grained Recognition, refer to the ACCV workshop competition in https://www.cvmart.net/race/10412/base.

Result

Show the result

LB A

LB B

Reproduce / 复现

For detailed self-supervised pretrain code, please refer to MMSelfSup. For detailed finetuning and inference code, please refer to this repo.

Description

Overview of Our Solution

Our Model

• ViT(MAE-pre-train) # Pretrained from MMSelfSup.
• Swin-v2(SimMIM-pre-train) # From MMCls-swin_transformer_v2.

**The architectures we use **

• ViT + CE-loss + post-LongTail-Adjusment
• ViT + SubCenterArcFaceWithAdvMargin(CE)
• Swin-B + SubCenterArcFaceWithAdvMargin(SoftMax-EQL)
• Swin-L + SubCenterArcFaceWithAdvMargin(SoftMAx-EQL)

bag of tricks paper and code

• MAE | Config
• Swinv2 | Config
• ArcFace | Code
• SubCenterArcFaceWithAdvMargin | Code
• Post-LT-adjusment | Code
• SoftMaxEQL | Code
• FlipTTA Code
• clean dataset
• self-emsemble: Uniform-model-soup | code
• pseudo | Code
• bagging-emsemble Code,
• post-process: re-distribute-label;

Used but no improvements

1. Using retrieval paradigm to solve this classification task;
2. Using EfficientNetv2 backbone.

Not used but worth to do

1. Try DiVE algorithm to improve performance in long tail dataset;
2. Use SimMIM to pre-train Swin-v2 on the competition dataset;
3. refine the re-distribute-label tool.