List your company

Audio-Visual Question Answering on MusicAVQA (Audio-centric target domain)

40.2Average ScoreVISCOP
27.82431.03734.2537.463Oct 15, 2025

Evaluation results

MethodLinks
VISCOPArchitecture=Universal Encoder (UE)2025.1040.211.9
UE + LLM (LoRA)Architecture=Universal Encoder (UE), Tuning=LoRA2025.1039.811.5
Base VLM2025.1028.3—