Identifying Same Persons From Temporally Synchronized Videos Taken by Multiple Wearable Cameras

Kang Zheng, Hao Guo, Xiaochuan Fan, Hongkai Yu, Song Wang; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2016, pp. 105-113

Abstract


Video-based human action recognition benefits from multiple cameras which can provide temporally synchronized, multi-view videos. Cross-video person identification, i.e., determining whether at a given time, persons tracked in different videos are the same person or not, is a key step to integrate multi-view information for collaborative action recognition. For fixed cameras, this step is relatively easy since they can be calibrated. In this paper, we study cross-video person identification for wearable cameras, which are constantly moving with the wearers. Specifically, we take tracked persons from different videos to be the same person if their 3D poses are the same, given these videos are synchronized. We adapt an existing algorithm to estimate the tracked person's 3D poses in each 2D video using motion-based features. Experiments show that, although 3D pose estimation is not perfect, it can still lead to better cross-video person identification than using appearance information.

Related Material


[pdf]
[bibtex]
@InProceedings{Zheng_2016_CVPR_Workshops,
author = {Zheng, Kang and Guo, Hao and Fan, Xiaochuan and Yu, Hongkai and Wang, Song},
title = {Identifying Same Persons From Temporally Synchronized Videos Taken by Multiple Wearable Cameras},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2016}
}