Efficient Visual Pointing for Embodied AI:Agent-Driven Data Synthesis, Cross-Block Attention, and Iterative Correction figure
AlphaXiv 中文概览(可滚动查看)