Learning a Discriminative Null Space for Person Re-identification

By Li Zhang, Tao A. Xiang, Shaogang Gong. Computer Vision and Pattern Recognition, 2016

668 citations recorded by the source corpus.

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.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/bb713964d37530310940aa5814cfb9b44ffad8b3

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

1. Yi Yang
2. Liang Zheng
3. Qi Tian
4. Shaogang Gong
5. Zhedong Zheng
6. Shiliang Zhang
7. Wanli Ouyang
8. Tao A. Xiang

In Computer Vision and Pattern Recognition

1. Yi Yang
2. Liang Zheng
3. Xiaogang Wang
4. Luc Van Gool
5. Dacheng Tao
6. Qi Tian
7. Wanli Ouyang
8. Ming-Hsuan Yang

Citations in the corpus

• Tracking the Untrackable: Learning to Track Multiple Cues with Long-Term Dependencies (2017, neutral): “DNS [90] 54.”
• Mask-Guided Contrastive Attention Model for Person Re-identification (2018, background): “In the past years, numerous of methods [47, 39, 25, 36, 26, 37, 43, 34] have been proposed to address this problem.”
• PROVID: Progressive and Multimodal Vehicle Reidentification for Large-Scale Urban Surveillance (2018, methodology): “Finally, the projection matrix W of the discriminative null space is learned by NFST on Φ(X) as in [38].”
• Scalable Person Re-identification on Supervised Smoothed Manifold (2017, methodology): “0 in rank-1 accuracy over Null [57] with detected bounding boxes.”
• Person Re-identification: Past, Present and Future (2016, methodology): “LOMO is later employed by [42], [43] and a similar set of features is used by Chen et al. [32].”
• Person Re-Identification by Camera Correlation Aware Feature Augmentation (2017, methodology): “Notable re-id learning models include PRDC [8], LADF [9], KISSME [13], PCCA [6], LFDA [10], XQDA [12], PSD [15], Metric Ensemble [14], DNS [39], SCSP [40], and so forth.”
• Adaptive Transfer Network for Cross-Domain Person Re-Identification (2019, neutral): “[20, 38, 43, 36].”
• Distilled Person Re-Identification: Towards a More Scalable System (2019, methodology): “A small sample learning ReID method DNS [63] was also compared, for which we used a PCB [47] model trained on MSMT17 as the best teacher model T1.”
• Person Re-Identification by Deep Joint Learning of Multi-Loss Classification (2017, background): “…matching distance metrics [Koestinger et al. , 2012; Xiong et al. , 2014; Zheng et al. , 2013; Wang et al. , 2014b; Paisitkri-angkrai et al. , 2015; Zhang et al. , 2016; Wang et al. , 2016b; Wang et al. , 2016c; Wang et al. , 2016d; Chen et al. , 2017b] or their combination in deep learning…”
• Beyond Triplet Loss: A Deep Quadruplet Network for Person Re-identification (2017, neutral): “40 NullReid [42] 58.”
• Learning Deep Context-Aware Features over Body and Latent Parts for Person Re-identification (2017, methodology): “Market1501: For the Market1501 dataset, several state-of-the-art methods are compared, including Bag of Words (BOW) [53], Weighted Approximate Rank Component Analysis (WARCA) [15], Discriminative Null Space (DNS) [47], Spatially Constrained Similarity function on Polynomial feature map (SCSP) [2], a”
• Applications of Generative Adversarial Networks (GANs): An Updated Review (2019, background): “[16, 117, 118, 138] and (2) developing a distinctive signature for representing a person under different camera environment having classification typically on cross-image representation [2, 111, 124].”
• Spindle Net: Person Re-identification with Human Body Region Guided Feature Decomposition and Fusion (2017, methodology): “The proposed Spindle Net is compared with several state-of-the-art ReID approaches in recent two years, including the metric learning method WARCA-χ2 [11], the null space semi-supervised learning method NFST [32], the deep neural network PersonNet [28], the Siamese CNN with learnable gate S-CNN [25]”
• Camera Style Adaptation for Person Re-identification (2017, methodology): “Examples in traditional approaches include KISSME [13], XQDA [17], DNS [34], etc. Examples in deep representation learning methods include IDE [37], SVDNet [24], TripletNet [8], etc. Camera 1 Camera 6 Comparing to previous methods, this paper resorts to an explicit strategy from the view of camera s”
• Attention-Aware Compositional Network for Person Re-identification (2018, methodology): “One set is poseirrelevant, which includes the null space semi-supervised learning method (NFST) [49], the domain guided dropout method (JSTL) [44], the deep transfer learning method (Transfer) [16], the Singular Vector Decomposition method (SVDNet) [39], the Online Instance Matching (OIM) method [45”