1 paper
Haoyuan Li, Qihang Cao, Tao Tang +6
Recent progress in spatial reasoning with Multimodal Large Language Models (MLLMs) increasingly leverages geometric priors from 3D encoders. However, most existing integration stra…