SPM-DomainNet-126

arxiv.org/abs/2505.24216

Code

Paper: https://ieeexplore.ieee.org/document/11084606

Shuffle PatchMix

This is the official implementation of the ICIP 2025 paper "Shuffle PatchMix Augmentation with Confidence-Margin Weighted Pseudo-Labels for Enhanced Source-Free Domain Adaptation", by Prasanna Reddy Pulakurthi, Majid Rabbani, Jamison Heard, Sohail A. Dianat, Celso M. de Melo, and Raghuveer Rao.

Shuffle PatchMix (SPM) is an augmentation technique that shuffles and blends image patches to generate diverse and challenging augmentations. It is combined with a novel reweighting strategy that prioritizes reliable pseudo-labels to mitigate label noise.

Installation

1. Clone this repository.

git clone https://github.com/PrasannaPulakurthi/SPM.git
cd SPM

1. Install requirements using Python 3.9.

conda create -n spm-env python=3.9
conda activate spm-env

1. The code is tested with Pytorch 1.7.1, CUDA 11.0.

pip install torch==1.7.1+cu110 torchvision==0.8.2+cu110 -f https://download.pytorch.org/whl/torch_stable.html

1. Please also make sure to install additional packages using the following command.

pip install -r requirements.txt

Datasets

The model was evaluated on three major benchmarks: PACS, VisDA-C, and DomainNet-126. Instructions for preparing these datasets can be found in the Github repository.