Omni-R1: Reinforcement Learning for Omnimodal Reasoning via Two-System Collaboration
Open paper
Benchmark
Model
Rank
Results
audio-visual-question-answering-on-omnibench
Omni-R1-7B
#3
Accuracy: 49.3
referring-video-object-segmentation-on-revos
Omni-R1-11B (Sa2VA-4B + SAM2 Hiera-L) (extra data)
#13
J&F: 58.9
Rank counts only results with a code link.