Publications
-
OmniBench: Towards The Future of Universal Omni-Language Models
NeurIPS 2025
-
OmniVideoBench: Towards Audio-Visual Understanding Evaluation for Omni MLLMs
ICLR 2026
-
MVU-Eval: Towards Multi-Video Understanding Evaluation for Multimodal LLMs
NeurIPS 2025
-
IF-VidCap: Can Video Caption Models Follow Instructions?
ICLR 2026
-
ViDiC: Video Difference Captioning
Preprint · arXiv 2025
-
Vibe AIGC: A New Paradigm for Content Generation via Agentic Orchestration
Preprint · arXiv 2026