hhtt2050/KANWave-Mamba

A Rice Leaf Disease Segmentation Method Based on Kolmogorov-Arnold Network and Wavelet-Enhanced State Space Model # KANWave-Mamba A Rice Leaf Disease Segmentation Method Based on Kolmogorov-Arnold Network and Wavelet-Enhanced State Space Model ## Data preparation ``` ├── data ├── ALDD ├── ImageSets ├── val.txt ├── train.txt ├── JPEGImages ├── im1.jpg ├── im2.jpg └── ... ├── Segmentation ├── im1.png ├── im2.png └── ... ├── RLDSD ├── ImageSets ├── val.txt ├── train.txt ├── JPEGImages ├── im1.jpg ├── im2.jpg └── ... ├── Segmentation ├── im1.png ├── im2.png └── ... ``` ### Prerequisites `conda` virtual environment is recommended. ```bash conda create -n kanwavemamba python=3.10 pip install -r requirements.txt ``` ## Training models. ``` cd code python train_KANWave_Mamba.py --disease RLDSD --name not_adv --root_path /KANWve-Mamba/data --batch_size 2 --max_iterations 80000 ```