Relaxed Pairwise Learned Metric for Person Re-identification

By Martin Hirzer, Peter M. Roth, Martin Köstinger, Horst Bischof. European Conference on Computer Vision, 2012

466 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/1a0a06e659eb075d414286d61bd36931770db799

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. Liang Zheng
2. Yi Yang
3. Qi Tian
4. Shaogang Gong
5. Shiliang Zhang
6. Xiatian Zhu
7. Tao A. Xiang
8. Wen Gao

In Computer Vision and Pattern Recognition

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

Citations in the corpus

• Multi-Task Learning with Low Rank Attribute Embedding for Multi-Camera Person Re-Identification (2018, methodology): “In this sense, our work is different from those algorithms performing distance metric learning [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16].”
• Person Re-Identification by Camera Correlation Aware Feature Augmentation (2017, methodology): “To this end, person re-id images are often represented by hand-crafted appearance pattern based features, designed and computed according to human domain knowledge [4], [12], [44], [45], [46], [50], [51], [52], [53], [54], [55].”
• Efficient PSD Constrained Asymmetric Metric Learning for Person Re-Identification (2015, methodology): “The metric learning approach has been applied to computer vision problems in recent years and shown to be effective [26, 7, 33, 9, 10, 13, 14].”
• Person re-identification by Local Maximal Occurrence representation and metric learning (2014, methodology): “Besides robust features, metric learning has been widely applied for person re-identification [43] [4] [11] [5] [49] [18] [14] [24].”
• Joint Learning of Single-Image and Cross-Image Representations for Person Re-identification (2016, methodology): “For the first category of methods, single-image representation (SIR) is first obtained using either hand-crafted feature [13, 14, 16, 28, 20, 17, 40, 36, 21, 22] or deep convolutional neural network (CNN) approaches [7, 37, 38], and then a distance measure together with a threshold is utilized to pr”
• Learning a Discriminative Null Space for Person Re-identification (2016, methodology): “The single shot version of the dataset is used in our experiments as in [15]: In each data split, 100 people with one image from each view are randomly chosen from the 200 present in both camera views for the training set, while the remaining 100 of View A are used as the probe set, and the remainin”
• Person Re-identification: Past, Present and Future (2016, background): “In recent years, Hirzer et al. [58] proposed relaxing the positivity constraint which provides a sufficient approximation for the matrix M with a much lower computational cost.”
• Re-id: Hunting Attributes in the Wild (2014, methodology): “Most feature representations can be used directly for re-identification by nearest neighbour (NN) matching or in conjunction with a model-based matching procedure such as metric learning [4, 14] or SVM [1, 28].”
• Clothing Attributes Assisted Person Reidentification (2015, methodology): “Recently, metric learning models have become popular in person reidentification, which include the probabilistic relative distance comparison model [7] and the relaxed Mahalanobis metric learning model [9].”
• Scalable Person Re-identification on Supervised Smoothed Manifold (2017, background): “Current research interests can be coarsely divided into two mainstreams: 1) those focus on designing robust visual descriptors [55, 13, 37, 26, 64] to accurately model the appearance of person; 2) those seek for a discriminative metric [65, 25, 54, 28, 18], under which instances of the same identity”
• Person Re-identification by Video Ranking (2014, methodology): “(2) Colour&LBP+DTW, HoGHoF+DTW: We applied Dynamic Time Warping [36] to compute the similarity of two sequences, using either Colour&LBP [20] or HoGHoF [27] as per-frame feature descriptor.”
• Similarity Learning with Spatial Constraints for Person Re-identification (2016, background): “The former type adopts the offthe-shelf distance metric such as Euclidean distance [10], Bhattacharyya distance[6], and covariance distance [22, 1], while the latter type tries to exploit the inherent invariance between image pairs [12, 27, 13, 24].”
• Unsupervised Salience Learning for Person Re-identification (2013, background): “Another direction is to learn task-specific distance functions with metric learning algorithms [29, 8, 24, 16].”
• Multi-scale Deep Learning Architectures for Person Re-identification (2017, unknown): “95 RPLM [17] 27.”
• Saliency Weighted Features for Person Re-identification (2014, methodology): “In Table 2, we also compare our method with RCCA [1], RPLM [21], PRDC [54], MCC [54], LAFT [28] and PCCA [42].”
• DeepReID: Deep Filter Pairing Neural Network for Person Re-identification (2014, background): “The purpose of person re-identification is to match pedestrians observed in non-overlapping camera views with visual features [13, 9, 35, 1, 7, 3, 51, 20, 19, 16, 29, 26, 24, 48, 2, 41].”
• Multi-Task Learning with Low Rank Attribute Embedding for Person Re-Identification (2015, methodology): “We first compare our approach with 8 completing methods for multi-short re-identification: Salience Matching (Salmatch) [47], Learning Mid-level Filters (LMF) [48], Multi-short Symmetry-driven Accumulation of Local Features (MS-SDALF) [12], Multi-short color with RankSVM (MS-color+RSVM) [41], Multi-”
• Deep Attributes Driven Multi-Camera Person Re-identification (2016, background): “Compared works that learn distance metrics for person ReID include RPML [9], PRDC [16], RSVM [49], Salmatch [45], LMF [46], PCCA [8], KISSME [12], kLFDA [13], KCCA [47],TSR [48], EPKFM [18],LOMO + XQDA [19],MRankPRDC [26], MRank-RSVM [26], RQDA [50], MLAPG [22] and CSL [21].”