双金字塔结构引导的多粒度行人重识别方法北大核心CSCD

针对杂乱场景下难以有效地提取行人关键信息和局部遮挡时全局特征方法失效的问题,提出了一种双金字塔结构引导的多粒度行人重识别(person re-identification,ReID)方法。首先在ResNet50中嵌入注意力金字塔,引导网络由粗到细依次挖掘不同粒度的特征,使网络更倾向于关注复杂环境中行人的显著区域;其次通过结构不对称的双重注意力特征金字塔分支(double attention feature pyramid branch,DFP branch)提取多尺度的行人特征,丰富特征的多样性,同时双重注意力机制可使分支从浅层信息中捕获高细粒度的局部特征;最后将粒度较粗的全局特征与多层级细粒度的局部特征融合,两种金字塔相互作用,以此获得更多具有鉴别性的多粒度特征,改善行人遮挡问题。在多个数据集上进行了实验,结果表明,各项评价指标均高于目前大多数主流模型,其中在DukeMTMC-reID数据集上,Rank-1、mAP和平均逆负处罚(mean inverse negative penalty,mINP)分别达到了91.6%、81.9%、48.1%。

Multi-granularity person re-identification method guided by double pyramid str ucture

Aiming at the problem that it is difficult to effectively extract the key information of pedestrians in the chaotic scene and the global feature method is invalid in the case of partial occlusion,a multi-granularity person re-identification (ReID) method guided by a dou ble pyramid structure is proposed.First,the attention pyramid in is embedded ResNet50 to guide the network to dig out features of different granularities from coarse to fine,making the network more inclined to focus on the significant areas of pedestrians in complex environments;secondly,the branch o f the double attention feature pyramid (DFP) with asymmetric structure is adopted.Multi-scale pedestrian fe atures are extracted to enrich the diversity of features.At the same time,the dual attention mechanism allows branches to capture finer-grained local features from shallow information;finally,the coa rser-grained global features are merged with multi-level and fine-grained local features,The two kinds of pyramids interact to retain more discriminative multi-granularity features to improve th e pedestrian occlusion problem.Experiments on multiple data sets have shown that the evaluation indica tors are higher than most current mainstream models.Among them,on the DukeMTMC-reID data set,Rank -1,mAP and mean inverse negative penalty (mINP) reached 91.6%,81.9% and 48.1%,respectively.