An open API service indexing awesome lists of open source software.

https://github.com/pris-cv/an-erudite-fgvc-model

Code release for Your “An Erudite Fine-Grained Visual Classification Model (CVPR 2023)"
https://github.com/pris-cv/an-erudite-fgvc-model

fine-grained fine-grained-classification fine-grained-recognition fine-grained-visual-categorization multiple-datasets pytorch

Last synced: about 1 year ago
JSON representation

Code release for Your “An Erudite Fine-Grained Visual Classification Model (CVPR 2023)"

Awesome Lists containing this project

README

          

# An-Erudite-FGVC-Model
Code release for “[An Erudite Fine-Grained Visual Classification Model](https://openaccess.thecvf.com/content/CVPR2023/papers/Chang_An_Erudite_Fine-Grained_Visual_Classification_Model_CVPR_2023_paper.pdf)" (CVPR 2023)

## Changelog
- 2023/04/18 upload the code.

## Requirements

- python 3.6
- PyTorch 1.7.1+cu110
- torchvision 0.8.2+cu110
- learn2learn 0.1.7

## Data
- Download datasets
- Extract them to `data/cars/`, `data/birds/` and `data/airs/`, respectively.
- Split the dataset into train and test folder, the index of each class should follow the Birds.xls, Air.xls, and Cars.xls

* e.g., CUB-200-2011 dataset
```
-/birds/train
└─── 001.Black_footed_Albatross
└─── Black_Footed_Albatross_0001_796111.jpg
└─── ...
└─── 002.Laysan_Albatross
└─── 003.Sooty_Albatross
└─── ...
-/birds/test
└─── ...
```

## Training
- `CUDA_VISIBLE_DEVICES=X python main.py` (Mix Cars and Flowers)

## Citation
If you find this paper useful in your research, please consider citing:
```
@InProceedings{Chang2023Erudite,
title={An Erudite Fine-Grained Visual Classification Model},
author={Chang, Dongliang and Tong, Yujun and Du, Ruoyi and Timothy, Hospedales and Song, Yi-Zhe and Ma, Zhanyu },
booktitle = {Computer Vision and Pattern Recognition},
year={2023}
}
```

## Contact
Thanks for your attention!
If you have any suggestion or question, you can leave a message here or contact us directly:
- changdongliang@bupt.edu.cn
- mazhanyu@bupt.edu.cn