跨模态行人重识别:Dynamic Dual-Attentive Aggregation Learningfor Visible-Infrared Person Re-Identification学习笔记
目录摘要方法模态内加权聚合(IWPA)跨模态图结构化注意力(CGSA)GraphConstructionGraphAttention动态对偶聚合学习试验论文链接:DynamicDual-AttentiveAggregationLearningforVisible-InfraredPersonRe-Identification摘要通过挖掘VI-ReID的模态内部分级和跨模态图级上下文线索,提出了一种新