AI RESEARCH
VGGSounder: Audio-Visual Evaluations for Foundation Models
arXiv CS.AI
•
ArXi:2508.08237v4 Announce Type: replace-cross The emergence of audio-visual foundation models underscores the importance of reliably assessing their multi-modal understanding. The VGGSound dataset is commonly used as a benchmark for evaluation audio-visual classification. However, our analysis identifies several limitations of VGGSound, including incomplete labelling, partially overlapping classes, and misaligned modalities. These lead to distorted evaluations of auditory and visual capabilities. To address these limitations, we