Anti-noise_FGVR

This repository provides a PyTorch implementation of the fine-grained vehicle recognition method, as proposed in my paper: Progressive Multi-Task Anti-Noise Learning and Distilling Frameworks for Fine-Grained Vehicle Recognition.

Figure 1: The target problem addressed by the proposed method

Figure 2: The proposed module

Environment

This source code was tested in the following environment:

Python = 3.8.13
PyTorch = 1.12.0
torchvision = 0.13.0
Ubuntu 20.04.6 LTS
NVIDIA GeForce RTX 3080 Ti

Pre-trained Models

The pre-trained models can be downloaded from this link.

Please save the downloaded models in the weightsFromCloud folder.

The xxxxx_Network.pth file was saved using torch.save(model, 'xxxxx_Network.pth').

The xxxxx_Weight.pth file was saved using torch.save(model.state_dict(), 'xxxxx_Weight.pth').

If you decide to register on InfiniCLOUD to download the model, I would appreciate it if you could kindly use my referral code XTQQJ during the process. This small gesture will be of great help to me.

Dependencies

• (1) Installation

Install Inplace-ABN following the instructions:

https://github.com/Alibaba-MIIL/TResNet/blob/master/requirements.txt

https://github.com/Alibaba-MIIL/TResNet/blob/master/INPLACE_ABN_TIPS.md

Install imgaug:

pip install imgaug

• (2) Download

Download the folder src from https://github.com/Alibaba-MIIL/TResNet, the folder vic from https://github.com/styler00dollar/pytorch-loss-functions, the folder example and Python file sam.py from https://github.com/davda54/sam, and save them as:

Anti-noise_FGVR
├── basic_conv.py
├── Inference_Stanford_Cars_ResNet50_Student.py
├── Inference_Stanford_Cars_ResNet50_Teacher.py
├── ...
├── src
├── vic
├── example
├── sam.py

Alternatively, you can simply download by running the following commands (note that subversion should be installed beforehand as sudo apt install subversion):

git clone https://github.com/Dichao-Liu/Anti-noise_FGVR.git
cd Anti-noise_FGVR
svn export https://github.com/Alibaba-MIIL/TResNet/branches/master/src
svn export https://github.com/davda54/sam/branches/main/example
svn export https://github.com/davda54/sam/branches/main/sam.py
svn export https://github.com/styler00dollar/pytorch-loss-functions/branches/main/vic

Dataset

• (1) Download the Stanford Cars dataset or other datasets mentioned in the paper, and organize the structure as follows:

dataset folder
├── 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) modify the path to the dataset folders.

Train

python Stanford_Cars_ResNet50_PMAL.py
python Stanford_Cars_ResNet50_Distillation.py

When training the student network, the --from_local option allows you to specify whether to use the teacher model downloaded from InfiniCLOUD or a model you have trained yourself using the provided code.

Inference

python Inference_Stanford_Cars_ResNet50_Teacher.py
python Inference_Stanford_Cars_ResNet50_Student.py

Bibtex

@ARTICLE{10623841,
  author={Liu, Dichao},
  journal={IEEE Transactions on Intelligent Transportation Systems}, 
  title={Progressive Multi-Task Anti-Noise Learning and Distilling Frameworks for Fine-Grained Vehicle Recognition}, 
  year={2024},
  volume={25},
  number={9},
  pages={10667-10678},
  keywords={Noise;Task analysis;Image recognition;Multitasking;Training;Noise measurement;Accuracy;Fine-grained vehicle recognition;intelligent transportation systems;ConvNets;object recognition},
  doi={10.1109/TITS.2024.3420151}
}