1 paper · 1 filter
Shiyun Xiong, Dongming Wu, Peiwen Sun +5
Recent embodied systems increasingly build on multimodal large language models (MLLMs) for their pretrained understanding and reasoning capabilities, but their perception of the ph…