CROC: Evaluating and Training T2I Metrics with Pseudo- and Human-Labeled Contrastive Robustness Checks
arXiv:2505.11314v2 Announce Type: replace-cross
Abstract: The assessment of evaluation metrics (meta-evaluation) is crucial for determining the suitability of existing metrics in text-to-image (T2I) generation tasks. Human-based meta-evaluation is cos…