DMRNet++: Learning Discriminative Features with Decoupled Networks and Enriched Pairs for One-Step Person Search

Authors: Chuchu Han, Zhedong Zheng, Kai Su, Dongdong Yu, Zehuan Yuan, Changxin Gao, Nong Sang, Yi Yang

Published in IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2022

Recommended citation: Chuchu Han, Zhedong Zheng, Kai Su, Dongdong Yu, Zehuan Yuan, Changxin Gao, Nong Sang, Yi Yang, "DMRNet++: Learning Discriminative Features with Decoupled Networks and Enriched Pairs for One-Step Person Search." IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022. DOI: 10.1109/TPAMI.2022.3221079
Download PDF: https://zdzheng.xyz/files/2022/Han_TPAMI22.pdf

Abstract: Person search aims at localizing and recognizing query persons from raw video frames, which is a combination of two sub-tasks,i.e., pedestrian detection and person re-identification. The dominant fashion is termed as the one-step person search that jointly optimizes detection and identification in a unified network, exhibiting higher efficiency . However, there remain major challenges: (i) conflicting objectives of multiple sub-tasks under the shared feature space, (ii) inconsistent memory bank caused by the limited batch size, (iii) underutilized unlabeled identities during the identification learning. T o address these issues, we develop an enhanced decoupled andmemory-reinforced network (DMRNet++). First, we simplify the standard tightly coupled pipelines and establish a task-decoupled framework (TDF). Second, we build a memory-reinforced mechanism (MRM), with a slow-moving average of the network to better encode the consistency of the memorized features. Third, considering the potential of unlabeled samples, we model the recognition process as semi-supervised learning. An unlabeled-aided contrastive loss (UCL) is developed to boost the identification feature learning by exploiting the aggregation of unlabeled identities. Experimentally , the proposed DMRNet++ obtains the mAP of 94.5\% and 52.1\% on CUHK-SYSU and PRW datasets, which exceeds most existing methods.

@article{han2022dmrnet++,
author = "Han, Chuchu and Zheng, Zhedong and Su, Kai and Yu, Dongdong and Yuan, Zehuan and Gao, Changxin and Sang, Nong and Yang, Yi",
title = "DMRNet++: Learning Discriminative Features with Decoupled Networks and Enriched Pairs for One-Step Person Search",
abstract = "Person search aims at localizing and recognizing query persons from raw video frames, which is a combination of two sub-tasks,i.e., pedestrian detection and person re-identification. The dominant fashion is termed as the one-step person search that jointly optimizes detection and identification in a unified network, exhibiting higher efficiency . However, there remain major challenges: (i) conflicting objectives of multiple sub-tasks under the shared feature space, (ii) inconsistent memory bank caused by the limited batch size, (iii) underutilized unlabeled identities during the identification learning. T o address these issues, we develop an enhanced decoupled andmemory-reinforced network (DMRNet++). First, we simplify the standard tightly coupled pipelines and establish a task-decoupled framework (TDF). Second, we build a memory-reinforced mechanism (MRM), with a slow-moving average of the network to better encode the consistency of the memorized features. Third, considering the potential of unlabeled samples, we model the recognition process as semi-supervised learning. An unlabeled-aided contrastive loss (UCL) is developed to boost the identification feature learning by exploiting the aggregation of unlabeled identities. Experimentally , the proposed DMRNet++ obtains the mAP of 94.5\\% and 52.1\\% on CUHK-SYSU and PRW datasets, which exceeds most existing methods.",
journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",
number = "01",
pages = "1--18",
year = "2022",
url = "https://zdzheng.xyz/files/2022/Han\_TPAMI22.pdf",
doi = "10.1109/TPAMI.2022.3221079",
publisher = "IEEE Computer Society" }