Learning a Discriminative Null Space for Person Re-identification
Top Cited Papers
- 1 June 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1239-1248
- https://doi.org/10.1109/cvpr.2016.139
Abstract
Most existing person re-identification (re-id) methods focus on learning the optimal distance metrics across camera views. Typically a person's appearance is represented using features of thousands of dimensions, whilst only hundreds of training samples are available due to the difficulties in collecting matched training images. With the number of training samples much smaller than the feature dimension, the existing methods thus face the classic small sample size (SSS) problem and have to resort to dimensionality reduction techniques and/or matrix regularisation, which lead to loss of discriminative power. In this work, we propose to overcome the SSS problem in re-id distance metric learning by matching people in a discriminative null space of the training data. In this null space, images of the same person are collapsed into a single point thus minimising the within-class scatter to the extreme and maximising the relative between-class separation simultaneously. Importantly, it has a fixed dimension, a closed-form solution and is very efficient to compute. Extensive experiments carried out on five person re-identification benchmarks including VIPeR, PRID2011, CUHK01, CUHK03 and Market1501 show that such a simple approach beats the state-of-the-art alternatives, often by a big margin.Keywords
This publication has 28 references indexed in Scilit:
- Matching People across Camera Views using Kernel Canonical Correlation AnalysisPublished by Association for Computing Machinery (ACM) ,2014
- People reidentification in surveillance and forensicsACM Computing Surveys, 2013
- Unsupervised Salience Learning for Person Re-identificationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2013
- Kernel Null Space Methods for Novelty DetectionPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2013
- Local Fisher Discriminant Analysis for Pedestrian Re-identificationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2013
- Person Re-identification by Descriptive and Discriminative ClassificationLecture Notes in Computer Science, 2011
- Information-theoretic metric learningPublished by Association for Computing Machinery (ACM) ,2007
- Null Foley–Sammon transformPattern Recognition, 2006
- Foley–Sammon Optimal Discriminant Vectors Using Kernel ApproachIEEE Transactions on Neural Networks, 2005
- A new LDA-based face recognition system which can solve the small sample size problemPattern Recognition, 2000