Cross-Granularity Feature Fusion

Code release for A Cross-Granularity Feature Fusion Method for Fine-Grained Image Recogniton

Requirement

python >= 3.7

PyTorch >= 1.3.1

torchvision >= 0.4.2

Training

1. Download datatsets for FGVC (e.g. CUB-200-2011, Standford Cars, FGVC-Aircraft, etc) and organize the structure as follows:

dataset
├── train
│   ├── class_001
|   |      ├── 1.jpg
|   |      ├── 2.jpg
|   |      └── ...
│   ├── class_002
|   |      ├── 1.jpg
|   |      ├── 2.jpg
|   |      └── ...
│   └── ...
└── test
    ├── class_001
    |      ├── 1.jpg
    |      ├── 2.jpg
    |      └── ...
    ├── class_002
    |      ├── 1.jpg
    |      ├── 2.jpg
    |      └── ...
    └── ...

2. Train from scratch with train.py.

If you find our code or paper useful to your research work, please consider citing our work using the following bibtex:

@article{wu2025cross, title={A cross-granularity feature fusion method for fine-grained image recognition}, author={Wu, Shan and Hu, Jun and Sun, Chen and Zhong, Fujin and Zhang, Qinghua and Wang, Guoyin}, journal={Applied Intelligence}, volume={55}, number={1}, pages={1--19}, year={2025}, publisher={Springer} }