WACVReview/WACV27-anon

MATCH-A: An Artificial Benchmark Dataset for Robust Image Matching in the GenAI Era is a large, fully synthetic dataset for training and evaluating robust image matching systems that link edited images to authentic sources. Why MATCH-A? The rapid spread of powerful editing and generative tools has made it easy to produce convincing fabrications. Integrity and fact-checking workflows rely on image matching, yet both everyday edits and AI-driven transformations can heavily degrade… See the full description on the dataset page: https://huggingface.co/datasets/WACVReview/WACV27-anon.