KD-TAO/LVOmniBench

LVOmniBench: Pioneering Long Audio-Video Understanding Evaluation for Omnimodal LLMs LVOmniBench is a new audio-visual understanding evaluation benchmark in long-form audio-video inputs. 🌟 🔥 News 2026.03.19 🌟 We are very proud to launch LVOmniBench, the pioneering comprehensive evaluation benchmark of OmniLLMs in Long Audio-Video Understanding Evaluation! ✨ LVOmniBench Introduction Recent advancements in omnimodal large language models… See the full description on the dataset page: https://huggingface.co/datasets/KD-TAO/LVOmniBench.