1 paper
Xiuwei Chen, Wentao Hu, Yongxin Wang +8
The integration of visual evidence has significantly enhanced the capabilities of large multimodal models. However, this integration predominantly relies on generating discrete out…