OmniCompass-AV
An evaluation track within OmniCompass aimed at omni-modal models, requiring models to answer using audio and visual evidence jointly. It comprises three task categories: temporal grounding, spatial grounding, and question answering.
Timeline (last 90 days)
OmniCompass is a holistic benchmark for omni-modal video localization and understanding, comprising two parallel evaluation tracks, OmniCompass-VL and OmniCompass-AV.
In OmniCompass-VL, question answering covers 5 perception categories and 3 reasoning categories; in OmniCompass-AV, question answering covers 3 perception categories and 2 reasoning categories.
All Facts (2)
OmniCompass is a holistic benchmark for omni-modal video localization and understanding, comprising two parallel evaluation tracks, OmniCompass-VL and OmniCompass-AV.
50%UnverifiedIn OmniCompass-VL, question answering covers 5 perception categories and 3 reasoning categories; in OmniCompass-AV, question answering covers 3 perception categories and 2 reasoning categories.
50%