DeepReID: Deep Filter Pairing Neural Network for Person Re-identification

By Wei Li, Rui Zhao, Tong Xiao, Xiaogang Wang. 2014

2,505 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

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

In Computer Vision and Pattern Recognition

1. Yi Yang
2. Luc Van Gool
3. Xiaogang Wang
4. Dacheng Tao
5. Ming-Hsuan Yang
6. Bernt Schiele
7. Qi Tian
8. Trevor Darrell

Citations in the corpus

• Faceless Person Recognition: Privacy Implications in Social Media (2016, background): “Like other areas of computer vision, the last years have seen a shift from classifiers based on handcrafted features and metric learning approaches [13, 14, 15, 16, 17, 18, 19] towards methods based on deep learning [20, 21, 22, 23, 24, 25, 26, 27].”
• Learning Deep Disentangled Embeddings with the F-Statistic Loss (2018, methodology): “For this task, we evaluate using two datasets—CUHK03 (Li et al., 2014) and Market1501 (Zheng et al.”
• In Defense of the Triplet Loss for Person Re-Identification (2017, background): “However, using pretrained networks also leads to a design lock-in, and does not allow for the exploration of new deep learning advances or different architectures.”
• Beyond frontal faces: Improving Person Recognition using multiple cues (2015, background): “[23] propose a deep network using pairs of people to encode photometric transformation.”
• Continual Representation Learning for Biometric Identification (2020, background): “Biometric identification [23,51], including face recognition [11, 37, 67] and person re-identification (re-id) [34, 66, 71], has achieved significant progress in the recent years due to the advances in modern learnable representations [8, 9, 11, 20, 34, 52, 60, 67] and emerging large datasets [17, 2”
• Person Recognition in Personal Photo Collections (2015, background): “Both settings have seen a recent shift from sophisticated classifiers based on hand-crafted features and metric learning approaches [20, 7, 5, 29, 26, 41, 1], towards methods based on deep learning [37, 36, 43, 33, 27, 38, 21].”
• TrackFormer: Multi-Object Tracking with Transformers (2021, methodology): “We indicated additional training Data: CH=CrowdHuman [44], PD=Parallel Domain [49] (synthetic), 6M=6 tracking datasets as in [64], JTA [13] (synthetic), M=Market1501 [65] and C=CUHK03 [26].”
• Deep feature learning with relative distance comparison for person re-identification (2015, methodology): “Here, we extend the triplet-based model to the person re-identification problem with an efficient learning algorithm and triplet generation scheme.”
• Reinforced Temporal Attention and Split-Rate Transfer for Depth-Based Person Re-identification (2017, background): “RGB-based person re-identification has progressed rapidly in recent years [40,89,1,82,79,73].”
• A Discriminatively Learned CNN Embedding for Person Reidentification (2016, methodology): “Recently, convolutional neural networks (CNN) have shown potential for learning state-of-the-art feature emb ddings or deep metrics [31, 12, 28, 23, 29, 35, 30].”
• Spindle Net: Person Re-identification with Human Body Region Guided Feature Decomposition and Fusion (2017, methodology): “5 Top1 accuracy on the CUHK03 dataset [13] , which is 10.”
• Scalable Person Re-identification: A Benchmark (2015, methodology): “We notice that the CUHK03 dataset [20] also uses the DPM detector, but the bboxes in CUHK03 are relatively good ones in terms of detector.”
• Multi-level Factorisation Net for Person Re-identification (2018, methodology): “MLFN is evaluated on three person Re-ID benchmarks, Market-1501 [55], CUHK03 [23] and DukeMTMCreID [56], and achieves state-of-the-art performance on all of them.”
• Learning a Discriminative Null Space for Person Re-identification (2016, methodology): “Datasets Five widely used datasets are selected for experiments, including the three largest benchmarks available (CUHK01, CUHK03, and Market1501).”
• Lifelong Person Re-Identification via Adaptive Knowledge Accumulation (2021, methodology): “Specifically, we selected five relatively large-scale datasets, CUHK03 (CU) [18], Market-1501 (MA) [52], MSMT17 V2 (MS) [40], DukeMTMC-ReID (DU) [54] and CUHK-SYSU ReID (SY) [43], and sampled 500 identities from each of their training sets to construct five training domains so that each domain has a”
• RGB-Infrared Cross-Modality Person Re-identification (2017, neutral): “) Datasets ID# images# cameras# RGB IR VIPER [7] 632 1,264 2 yes no iLIDS [54] 119 476 2 yes no CAVIAR [5] 72 610 2 yes no PRID2011 [11] 200 971 2 yes no CUHK01 [19] 972 1,942 2 yes no SYSU [8] 502 24,448 2 yes no CUHK03 [20] 1467 13,164 6 yes no Market [53] 1501 32,668 6 yes no MARS [52] 1261 1,191”
• Scalable Person Re-identification on Supervised Smoothed Manifold (2017, methodology): “In on-line stage, the extra indexing time training data provided by CUHK03, the extra indexing cost is 0 .”
• Beyond Part Models: Person Retrieval with Refined Part Pooling (2017, neutral): “DukeMTMC-reID CUHK03 rank-1 mAP rank-1 mAP BoW+kissme [39] 25 under the single-query mode without re-ranking.”