A Strong Baseline and Batch Normalization Neck for Deep Person Re-Identification

作     者：Luo, Hao Jiang, Wei Gu, Youzhi Liu, Fuxu Liao, Xingyu Lai, Shenqi Gu, Jianyang 

作者机构：Zhejiang Univ Coll Control Sci & Enginneering State Key Lab Ind Control Technol Hangzhou 310027 Peoples R China Ping An Technol Shenzhen 518000 Peoples R China Chinese Acad Sci Beijing 100190 Peoples R China Xi An Jiao Tong Univ Xian 710049 Peoples R China 

出 版 物：《IEEE TRANSACTIONS ON MULTIMEDIA》 (IEEE多媒体汇刊)

年 卷 期：2020年第22卷第10期

页      面：2597-2609页

核心收录：

学科分类：0810[工学-信息与通信工程] 0808[工学-电气工程] 08[工学] 0835[工学-软件工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：National Natural Science Foundation of China Science Foundation of Chinese Aerospace Industry [JCKY2018204B053] Autonomous Research Project of the State Key Laboratory of Industrial Control Technology, China [ICT1917] 

主　　题：Person ReID baseline tricks BNNeck deep learning 

摘      要：This study proposes a simple but strong baseline for deep person re-identification (ReID). Deep person ReID has achieved great progress and high performance in recent years. However, many state-of-the-art methods design complex network structures and concatenate multi-branch features. In the literature, some effective training tricks briefly appear in several papers or source codes. The present study collects and evaluates these effective training tricks in person ReID. By combining these tricks, the model achieves 94.5% rank-1 and 85.9% mean average precision on Market1501 with only using the global features of ResNet50. The performance surpasses all existing global- and part-based baselines in person ReID. We propose a novel neck structure named as batch normalization neck (BNNeck). BNNeck adds a batch normalization layer after global pooling layer to separate metric and classification losses into two different feature spaces because we observe they are inconsistent in one embedding space. Extended experiments show that BNNeck can boost the baseline, and our baseline can improve the performance of existing state-of-the-art methods.