On March 11, the Tongyi Lab team at Alibaba Cloud announced the open sourcing of the R1-Omni model. This innovative model seamlessly integrates reinforcement learning with verifiable reward mechanisms, resulting in a substantial enhancement of multimodal emotion recognition capabilities. This milestone represents a significant advancement in the evolution of full-modal models. Experimental results have demonstrated R1-Omni's exceptional performance across multiple test sets.
