Learning Structured Visual Compositional Representations for Weakly Supervised Referring Expression Comprehension figure
AlphaXiv 中文概览(可滚动查看)