AI RESEARCH

NeXT-IMDL: Build Benchmark for NeXT-Generation Image Manipulation Detection & Localization

arXiv CS.CV

ArXi:2512.23374v2 Announce Type: replace The accessibility surge and abuse risks of user-friendly image editing models have created an urgent need for generalizable, up-to-date methods for Image Manipulation Detection and Localization (IMDL). Current IMDL research typically uses cross-dataset evaluation, where models trained on one benchmark are tested on others. However, this simplified evaluation approach conceals the fragility of existing methods when handling diverse AI-generated content, leading to misleading impressions of progress.