Multi-Modal Context Propagation for Person Re-Identification With Wireless Positioning

Existing person re-identification methods mainly rely on the visual appearance captured by cameras for identity matching. However, dueto the sensitivity of visual data to occlusion, blur, clothing change, etc. , existing methods struggle to distinguish pedestrians in challenging scenarios. Inspired...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on multimedia 2022, Vol.24, p.3060-3073
Main Authors: Liu, Yiheng, Zhou, Wengang, Xi, Mao, Shen, Sanjing, Li, Houqiang
Format: Article
Language:English
Subjects:
Cameras
Cell phones
Cellular communication
Context
Identification methods
multi-modal recognition
Occlusion
Pedestrians
Person re-identification
Propagation
Trajectory
unsupervised cross-domain learning
Video sequences
Visual signals
Visualization
Wireless communication
Wireless fidelity
Wireless networks
wireless positioning
Wireless sensor networks
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Existing person re-identification methods mainly rely on the visual appearance captured by cameras for identity matching. However, dueto the sensitivity of visual data to occlusion, blur, clothing change, etc. , existing methods struggle to distinguish pedestrians in challenging scenarios. Inspired by the fact that most pedestrians carry around smart wireless devices, e.g. mobile phones that can be sensed by WiFi or cellular networks as wireless positioning signals, we propose to exploit the free yet informative wireless signals to assist person re-identification. It is well recognized that wireless signals are robust to visual noises mentioned above, which perform as a good complement to the visual data. To make full use of these multi-modal clues for person re-identification, we propose a multi-modal context propagation framework MCPF that contains a recurrent context propagation module RCPM and an unsupervised multi-modal cross-domain method UMM-ReID. RCPM enables context information to be continuously propagated and fused between visual data and wireless data. UMM-ReID utilizes wireless signals to constrain the estimation of pseudo labels. We contribute a new wireless positioning person re-identification dataset WP-ReID to evaluate our approach. Extensive experiments demonstrate the effectiveness of the proposed method. Benefiting from the collaboration of RCPM and UMM-ReID, the proposed framework MCPF achieves a significant performance improvement over existing methods.
ISSN:1520-9210
1941-0077
DOI:10.1109/TMM.2021.3092579