Multi-Scale Triplet CNN for Person Re-Identification
Multi-Scale Triplet CNN for Person Re-Identification
复制标题
DOI：
10.1145/2964284.2967209
复制
发表时间：
2016-10
期刊：
影响因子：
--
通讯作者：
Jiawei Liu;Zhengjun Zha;Q. Tian;Dong Liu;Ting Yao;Q. Ling;Tao Mei
中科院分区：
文献类型：
--
作者：
Jiawei Liu;Zhengjun Zha;Q. Tian;Dong Liu;Ting Yao;Q. Ling;Tao Mei
人员重识别旨在跨非重叠的多摄像头网络识别特定人员。这是自动视频监控中的一项基本且具有挑战性的任务。现有的大多数研究主要依赖于手工制作的特征，导致性能不理想。在本文中，我们提出了一种多尺度三元组卷积神经网络，它可以捕获不同尺度的人的视觉外观。我们建议通过大量样本三元组上的比较相似性损失来优化网络参数，解决行人重识别中训练集较小的问题。特别是，我们设计了一个由深层和浅层神经网络组成的统一的多尺度网络架构，旨在学习复杂条件下人员重新识别的鲁棒有效的特征。对现实世界 Market-1501 数据集的广泛评估证明了所提出方法的有效性。
Person re-identification aims at identifying a certain person across non-overlapping multi-camera networks. It is a fundamental and challenging task in automated video surveillance. Most existing researches mainly rely on hand-crafted features, resulting in unsatisfactory performance. In this paper, we propose a multi-scale triplet convolutional neural network which captures visual appearance of a person at various scales. We propose to optimize the network parameters by a comparative similarity loss on massive sample triplets, addressing the problem of small training set in person re-identification. In particular, we design a unified multi-scale network architecture consisting of both deep and shallow neural networks, towards learning robust and effective features for person re-identification under complex conditions. Extensive evaluation on the real-world Market-1501 dataset have demonstrated the effectiveness of the proposed approach.