AI RESEARCH
FairJudge: Abstention-Aware Multimodal Judges for Fairness and Alignment Evaluation in Text-to-Image Models
arXiv CS.LG
•
ArXi:2510.22827v3 Announce Type: replace-cross Evaluating text-to-image (T2I) systems requires judging not only whether an image matches a prompt, but also whether socially salient attributes are represented faithfully and without uned inference. Existing automated evaluators typically rely on face-centric recognizers or contrastive image--text similarity, which provide limited diagnostic feedback and often force predictions even when visual evidence is ambiguous or absent.