Robustness Evaluation of Image Inpainting Techniques

This repository provides the codebase for evaluating six image inpainting methods—Median, k-Nearest Neighbors (kNN), Matrix Completion, Variational Autoencoder with Weighted Loss (VAE-WL), Masked Autoencoder Vision Transformer (MAE-ViT), and Masked Autoencoder Vision Transformer with GAN Loss (MAE-ViT+GAN)—across three mammography datasets: INBreast, MIAS, and a stratified 1,000-image subset of VinDr-Mammo.

An example of inpainted images produced by the evaluated methods is shown below:

The results reported in this repository are described in a paper submitted to the 28th International Conference on Pattern Recognition (ICPR 2026), to be held in Lyon, France. Additionally, the supplementary material can be viewed here

Getting Started

We recommend creating a virtual environment before running the experiments:

python -m venv env
source env/bin/activate  # On Linux/macOS
.\env\Scripts\activate   # On Windows

To install the required dependencies, run:

pip install -r requirements.txt

Acknowledgments

This study was financed, in part, by the São Paulo Research Foundation (FAPESP), Brasil. Process Numbers 2021/06870-3 and 2024/23791-8. This work was also financed through national funds by FCT - Fundação para a Ciência e a Tecnologia, I.P., in the framework of the Project UIDB/00326/2025 and UIDP/00326/2025. Additionally, it was supported by the Portuguese Recovery and Resilience Plan (PRR) through project C645008882-00000055-Center for Responsable AI.