AVQA benchmarks
Emerging6papers using it
2023first seen
The 'AVQA benchmarks' is a dataset used to evaluate audio-visual question answering systems by providing modality-preference labels that guide the reasoning process in distinguishing and integrating audio and visual information.
Papers using AVQA benchmarks (6)
- OmniEncoder: See, Hear, and Feel Continuous Motion Like Humans With One EncoderSeparate First, Fuse Later: Mitigating Cross-Modal Interference in Audio-Visual LLMs Reasoning with Modality-Specific Chain-of-ThoughtNot All Modalities Are Equal: Instruction-Aware Gating for Multimodal VideosCAE-AV: Improving Audio-Visual Learning via Cross-modal Interactive EnrichmentSightSound-R1: Cross-Modal Reasoning Distillation from Vision to Audio Language ModelsONE-PEACE: Exploring One General Representation Model Toward Unlimited
Modalities