Re-Write CMAL-Net for 43999 Project

Source

Original Paper | CMAL-Net | TResNet

About

This is some my-self rewrite the code of CMAL for Project Sustainability Energy Fine-Grained Image Classification. I just do some rewriting (to more understand the code) and change a litte bit deprecated code and also make a litte bit modular.

The focus rewrite is the Model where using Pre-trained of TResNet-L.

Environment Requirements

This repository actualy running on Kaggle / Colab Notebook. To Ensure that your environment meets the following specifications:

• Python: 3.10.12
• PyTorch: 2.0.0
• Torchvision: 0.15.1
• Ubuntu: 22.04.3 LTS
• CUDA: 11.4 (or newer)

Dependencies

To ensure completeness, refer to the original source here

For Dependecies is 2:

1. Inplace-ABN

For the Inplace-ABN dependency, follow the installation instructions outlined here.

Install Inplace-ABN using pip:

pip install inplace-abn

Note: Ensure that CUDA is installed to enable successful installation.

Dataset

Download your dataset and organize the structure to follow like this:

dataset/
├── train/
│   ├── class_1/
│   │   ├── 01.jpg
│   │   ├── 02.jpg
│   │   └── ...
│   ├── class_2/
│   │   ├── 01.jpg
│   │   ├── 02.jpg
│   │   └── ...
│   └── ...
└── test/
    ├── class_1/
    │   ├── 01.jpg
    │   ├── 02.jpg
    │   └── ...
    ├── class_2/
    │   ├── 01.jpg
    │   ├── 02.jpg
    │   └── ...
    └── ...