This repository is for Kaggle Competetion Plant Pathology 2020 - FGVC7. And our rank is 5/1115.
| model | score(AUC) |
|---|---|
| SeResNet50 | 96.0 |
| SeResNet101 | 96.7 |
| EffcientNet b4 | 97.5 |
| EffcientNet b6 | 97.7 |
| ResNest101 | 97.3 |
| ResNest200 | 97.3 |
| WSDAN* | 97.0 |
| merge** | 98.6 |
*. WSDAN model is from pytorch implement of See Better Before Looking Closer: Weakly Supervised Data Augmentation Network for Fine-Grained Visual Classification
**. merge model is the merge of EffcientNet b4, EffcientNet b6(2 models), EffcientNet b5, EffcientNet b7, ResNest101,
python train.py --net $YOURNET --loss allYOURNET should be in b0~b7, seresnet50, seresnet101,resnest101, resnest200, resnest269
python eval_tta.py --net $YOURNET