Omni Interaction Agent Technical Report

Benchmark Model Rank Results
audio-question-answering-on-openaudiobench-llamaquestionsGander#5Accuracy: 75.6
audio-visual-question-answering-on-dailyomniGander#5Accuracy: 78.53
audio-visual-question-answering-on-worldsenseGander#7Accuracy: 49.62