# MSENet

**Repository Path**: LHR_97/MSENet

## Basic Information

- **Project Name**: MSENet
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-09-06
- **Last Updated**: 2025-09-06

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Enhancing Few-Shot Image Classification through Learnable Multi-Scale Embedding and Attention Mechanisms
Implementation of a Few-Shot Image Classification Model based on the Prototypical Network Model and Tested on the MiniImagenet and FC100 Datasets.

For more information, check out our paper on [[arXiv](https://arxiv.org/abs/2409.07989)], [[paperswithcode](https://paperswithcode.com/paper/enhancing-few-shot-image-classification)].

## Model

Our model consists of the following components:
1. We extracted five feature maps from backbone in order to capture both global and task specific features
2. We employ a self-attention mechanism for each feature map obtained from every stage in order to capture more valuable information
3. We incorporate learnable weights at each stage.
4. We propose a novel few-shot classification. We have
significantly improved the accuracy on the MiniImageNet and FC100 datasets.

The final model architecture is as follows:

![Architecture of model](assets/finalmodel.png)

The mapper architecture is as follows:

![Architecture of mapper](assets/attention-module.png)

You can study the model in more detail from this [PDF](finalreport.pdf).

## How to run
For the 5-way 5-shot:
```bash
python train.py --max-epoch 200 --save-epoch 20 --shot 5 --query 10 --train-way 30 --test-way 5 --save-path ./save/proto-5-change --gpu 0
```

For the 5-way 1-shot:
```bash
python train.py --max-epoch 200 --save-epoch 20 --shot 1 --query 10 --train-way 20 --test-way 5 --save-path ./save/proto-1-change --gpu 0
```

## Comparation

![MiniImageNet](assets/table1.JPG)

![FC100](assets/table2.JPG)

![CUB](assets/table3.JPG)

## Citation
If you use this repository in your work, please cite the following paper:
```bibtex
@article{askari2025enhancing,
  title={Enhancing few-shot image classification through learnable multi-scale embedding and attention mechanisms},
  author={Askari, Fatemeh and Fateh, Amirreza and Mohammadi, Mohammad Reza},
  journal={Neural Networks},
  volume={187},
  pages={107339},
  year={2025},
  publisher={Elsevier}
}