DifFSS: Diffusion Model for Few-Shot Semantic Segmentation

This is the implementation of the paper "DifFSS: Diffusion Model for Few-Shot Semantic Segmentation".

1. Preparation of the auxiliary support images

Please refer to our ControlNet4FSS pipeline for the pre-generattion of auxiliary support images.

2. DifFSS on FSS models

The DifFSS method can be easily applied to existing FSS models, see one of our implementations on BAM here.