Uni-MoE-2.0-Omni: Scaling Language-Centric Omnimodal Large Model with Advanced MoE, Training and Data

Benchmark Model Rank Results
audio-visual-question-answering-on-omnibenchUni-MoE-2.0-Omni#4Accuracy: 47.1
audio-visual-question-answering-on-omnivideobenchUni-MoE-2.0-Omni#8Accuracy: 35.1
audio-visual-question-answering-on-worldsenseUni-MoE-2.0-Omni#15Accuracy: 44.7