【资源】行人重识别(Person ReID)相关资源

行人重识别(Person Re-identification也称行人再识别,简称为ReID,是利用计算机视觉技术判断图像或者视频序列中是否存在特定行人的技术。广泛被认为是一个图像检索的子问题。给定一个监控行人图像,检索跨设备下的该行人图像。本文分享一个行人重识别的资源列表,含综述、论文(CVPR2019,监督/无监督等)、数据集等。

更多Awsome Github资源请关注:【Awsome】GitHub 资源汇总


Awesome Person Re-identification (Person ReID) Awesome


Table of Contents



1. Statistics

Conference Link #Total Unsupervised reID Supervised reID Person search Others Datasets
CVPR2019 will be updated soon 29 ? ? ? ? ?
ECCV2018 Click 19 4 8 4 3 0
CVPR2018 Click 31 5 23 0 2 1
ICCV2017 Click 16 7 6 1 2 0
CVPR2017 Click 16 2 9 1 3 1



2. Survey

1) "Person Re-identification: Past, Present and Future", arXiv 2016 [paper]

2) "A survey of approaches and trends in person re-identification", Image and Vision Computing 2014 [paper]

3) "Appearance Descriptors for Person Re-identification: a Comprehensive Review", arXiv 2013 [paper]




3. CVPR 2019

(not categorized yet)

1) "Joint Discriminative and Generative Learning for Person Re-identification" [paper]

2) "Densely Semantically Aligned Person Re-Identification" [paper]

3) "Generalizable Person Re-identification by Domain-Invariant Mapping Network" [paper]

4) "Re-Identification with Consistent Attentive Siamese Networks" [paper]

5) "Distilled Person Re-identification: Towards a More Scalable System" [paper]

6) "Weakly Supervised Person Re-Identification" [paper]

7) "Patch Based Discriminative Feature Learning for Unsupervised Person Re-identification" [paper]

8) "Unsupervised Person Re-identification by Soft Multilabel Learning" [paper] [github]

9) "Invariance Matters: Exemplar Memory for Domain Adaptive Person Re-identification" [paper] [github]

10) "Re-ranking via Metric Fusion for Object Retrieval and Person Re-identification" [paper]

11) "Progressive Pose Attention Transfer for Person Image Generation" [paper] [github]

12) "Unsupervised Person Image Generation with Semantic Parsing Transformation" [paper] [github]

13) "Learning to Reduce Dual-level Discrepancy for Infrared-Visible Person Re-identification" [paper]

14) "Text Guided Person Image Synthesis" [paper]

15) "Re-Identification Supervised 3D Texture Generation" [paper]

16) "Learning Context Graph for Person Search" [paper] [github]

17) "Query-guided End-to-End Person Search" [paper]

18) "Multi-person Articulated Tracking with Spatial and Temporal Embeddings" [paper]

19) "Dissecting Person Re-identification from the Viewpoint of Viewpoint" [paper]

20) "Towards Rich Feature Discovery with Class Activation Maps Augmentation for Person Re-Identification"

21) "AANet: Attribute Attentio Network for Person Re-Identification"

22) "VRSTC: Occlusion-Free Video Person Re-Identification"

23) "Adaptive Transfer Network for Cross-Domain Person Re-Identification"

24) "Pyramidal Person Re-IDentification via Multi-Loss Dynamic Training" [paper]

25) "Interaction-and-Aggregation Network for Person Re-identification"

26) "Skin-based identification from multispectral image data using CNNs"

27) "Feature Distance Adversarial Network for Vehicle Re-Identification"

28) "Part-regularized Near-Duplicate Vehicle Re-identification" [paper]

29) "CityFlow: A City-Scale Benchmark for Multi-Target Multi-Camera Vehicle Tracking and Re-Identification" [paper]




4. Unsupervised Person Re-ID

(+ Transfer learning / Semi-supervised learning / Weakly-supervised learning)


[ECCV2018]

1) "Domain Adaptation through Synthesis for Unsupervised Person Re-identification" [paper]

  • Performance
    • Rank-1 (Market1501) : 65.7 (Single query)

      2) "Unsupervised Person Re-identification by Deep Learning Tracklet Association" [paper]

  • Performance
    • Rank-1 (Market1501) : 63.7 (ResNet-50. Images resized to 256x128.)
  • Summary
    • Sparse space-time tracklet (SSTT) sampling
    • 1) Sparse temporal sampling
    • 2) Sparse spatial sampling
    • Tracklet association unsupervised deep learning (TAUDL) in an end-to-end manner.
    • 1) Per-camera tracklet discrimination learning
    • 2) Cross-camera tracklet association learning

      3) "Generalizing A Person Retrieval Model Hetero- and Homogeneously" [paper] [Github]

  • Performance
    • Rank-1 (Market1501) : 62.2(Single query, source domain : DukeMTMC) / 56.8(Single query, source domain : CUHK03)

      4) "Robust Anchor Embedding for Unsupervised Video Person Re-Identification in the Wild" [paper]


[CVPR2018]

1) "Unsupervised Cross-dataset Person Re-identification by Transfer Learning of Spatial-Temporal Patterns" [paper] [Github]

  • Performance
    • Rank-1 (Market1501) : 60.7(source domain : CUHK01) / 59.2(source domain : VIPeR) / 58.2(source domain : GRID)

      2) "Transferable Joint Attribute-Identity Deep Learning for Unsupervised Person Re-Identification" [paper]

  • Performance
    • Rank-1 (Market1501) : 58.2(source domain : DukeMTMC)

      3) "Image-Image Domain Adaptation with Preserved Self-Similarity and Domain-Dissimilarity for Person Re-identification" [paper] [Github]

  • Rank-1 (Market1501) : 58.1

    4) "Disentangled Person Image Generation" [paper]

  • Performance
    • Rank-1 (Market1501) : 35.5

      5) "Exploit the Unknown Gradually: One-Shot Video-Based Person Re-Identification by Stepwise Learning" [paper] [Github] [Homepage]


[ICCV2017]

1) "Cross-view Asymmetric Metric Learning for Unsupervised Person Re-identification" [paper] [Github]

  • Performance
    • Rank-1 (Market1501) : 54.5(Multiple query)

      2) "Efficient Online Local Metric Adaptation via Negative Samples for Person Re-Identification" [paper]

  • Performance
    • Rank-1 (Market1501) : 51.5(Multiple query) / 40.9(Single query)

      3) "SHaPE: A Novel Graph Theoretic Algorithm for Making Consensus-based Decisions in Person Re-identification Systems" [paper]

      4) "Stepwise Metric Promotion for Unsupervised Video Person Re-identification" [paper]

      5) "Group Re-Identification via Unsupervised Transfer of Sparse Features Encoding" [paper]

      6) "Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitro" [paper] [Github]

      7) "Dynamic Label Graph Matching for Unsupervised Video Re-Identification" [paper] [Github]


[CVPR2017]

1) "One-Shot Metric Learning for Person Re-Identification" [paper]

2) "Unsupervised Adaptive Re-Identification in Open World Dynamic Camera Networks" [paper]


[arXiv]

1) "Unsupervised Person Re-identification: Clustering and Fine-tuning" [paper] [Github]

  • Performance
    • Rank-1 (Market1501) : 41.9 (Single query, source domain : CUHK03)



5. Supervised Person Re-ID

[ECCV2018]

1) "Maximum Margin Metric Learning Over Discriminative Nullspace for Person Re-identification" [paper]

2) "Person Re-identification with Deep Similarity-Guided Graph Neural Network" [paper]

3) "Pose-Normalized Image Generation for Person Re-identification" [paper]

4) "Improving Deep Visual Representation for Person Re-identification by Global and Local Image-language Association" [paper]

5) "Hard-Aware Point-to-Set Deep Metric for Person Re-identification" [paper]

6) "Part-Aligned Bilinear Representations for Person Re-Identification" [paper] [Github]

  • Summary
    • Two stream network (Appearance map extractor, Part map extractor-OpenPose)
    • Aggregator : bilinear pooling (better than concat + Ave. pool + linear)

      7) "Mancs: A Multi-task Attentional Network with Curriculum Sampling for Person Re-identification" [paper]

      8) "Beyond Part Models: Person Retrieval with Refined Part Pooling (and A Strong Convolutional Baseline)" [paper]


[CVPR2018]

1) "Diversity Regularized Spatiotemporal Attention for Video-based Person Re-identification" [paper]

2) "A Pose-Sensitive Embedding for Person Re-Identification with Expanded Cross Neighborhood Re-Ranking" [paper]

3) "Human Semantic Parsing for Person Re-identification" [paper]

4) "Video Person Re-identification with Competitive Snippet-similarity Aggregation and Co-attentive Snippet Embedding" [paper]

5) "Mask-guided Contrastive Attention Model for Person Re-Identification" [paper]

6) "Person Re-identification with Cascaded Pairwise Convolutions" [paper]

7) "Multi-Level Factorisation Net for Person Re-Identification" [paper]

8) "Attention-Aware Compositional Network for Person Re-identification" [paper]

9) "Deep Group-shuffling Random Walk for Person Re-identification" [paper]

10) "Harmonious Attention Network for Person Re-Identification" [paper]

11) "Efficient and Deep Person Re-Identification using Multi-Level Similarity" [paper]

12) "Pose Transferrable Person Re-Identification" [paper]

13) "Adversarially Occluded Samples for Person Re-identification" [paper]

14) "Camera Style Adaptation for Person Re-identification" [paper]

15) "Dual Attention Matching Network for Context-Aware Feature Sequence based Person Re-Identification" [paper]

16) "Easy Identification from Better Constraints: Multi-Shot Person Re-Identification from Reference Constraints" [paper]

17) "Eliminating Background-bias for Robust Person Re-identification" [paper]

18) "Features for Multi-Target Multi-Camera Tracking and Re-Identification" [paper]

19) "Multi-shot Pedestrian Re-identification via Sequential Decision Making" [paper]

20) "End-to-End Deep Kronecker-Product Matching for Person Re-identification" [paper]

21) "Deep Spatial Feature Reconstruction for Partial Person Re-identification: Alignment-free Approach" [paper]

22) "Resource Aware Person Re-identification across Multiple Resolutions" [paper]

23) "Group Consistent Similarity Learning via Deep CRF for Person Re-Identification" [paper]


[ICCV2017]

1) "A Two Stream Siamese Convolutional Neural Network For Person Re-Identification" [paper]

2) "Learning View-Invariant Features for Person Identification in Temporally Synchronized Videos Taken by Wearable Cameras" [paper]

3) "Deeply-Learned Part-Aligned Representations for Person Re-Identification" [paper]

4) "Pose-driven Deep Convolutional Model for Person Re-identification" [paper]

5) "Jointly Attentive Spatial-Temporal Pooling Networks for Video-based Person Re-Identification" [paper]

6) "Multi-scale Deep Learning Architectures for Person Re-identification" [paper]


[CVPR2017]

1) "Learning Deep Context-Aware Features Over Body and Latent Parts for Person Re-Identification" [paper]

2) "Beyond Triplet Loss: A Deep Quadruplet Network for Person Re-Identification" [paper]

3) "Spindle Net: Person Re-Identification With Human Body Region Guided Feature Decomposition and Fusion" [paper]

4) "Re-Ranking Person Re-Identification With k-Reciprocal Encoding" [paper]

5) "Scalable Person Re-Identification on Supervised Smoothed Manifold" [paper]

6) "Point to Set Similarity Based Deep Feature Learning for Person Re-Identification" [paper]

7) "Fast Person Re-Identification via Cross-Camera Semantic Binary Transformation" [paper]

8) "See the Forest for the Trees: Joint Spatial and Temporal Recurrent Neural Networks for Video-Based Person Re-Identification" [paper]

9) "Consistent-Aware Deep Learning for Person Re-Identification in a Camera Network" [paper]




6. Person Search

[ECCV2018]

1) "RCAA: Relational Context-Aware Agents for Person Search" [paper]

2) "Person Search in Videos with One Portrait Through Visual and Temporal Links" [paper]

3) "Person Search by Multi-Scale Matching" [paper]

4) "Person Search via A Mask-Guided Two-Stream CNN Model" [paper]


[ICCV2017]

1) "Neural Person Search Machines" [paper]


[CVPR2017]

1) "Person Search with Natural Language Description" [paper]




7. Others

[ECCV2018]

1) "Integrating Egocentric Videos in Top-view Surveillance Videos: Joint Identification and Temporal Alignment" [paper]

2) "Reinforced Temporal Attention and Split-Rate Transfer for Depth-Based Person Re-Identification" [paper]

3) "Adversarial Open-World Person Re-Identification" [paper]


[CVPR2018]

1) "Viewpoint-aware Attentive Multi-view Inference for Vehicle Re-identification" [paper]

2) "Exploiting Transitivity for Learning Person Re-identification Models on a Budget" [paper]


[ICCV2017]

1) "Orientation Invariant Feature Embedding and Spatial Temporal Regularization for Vehicle Re-identification" [paper]

2) "RGB-Infrared Cross-Modality Person Re-Identification" [paper]


[CVPR2017]

1) "Joint Detection and Identification Feature Learning for Person Search" [paper]

2) "Multiple People Tracking by Lifted Multicut and Person Re-Identification" [paper]

3) "Pose-Aware Person Recognition" [paper]



8. Datasets

[CVPR2018]

1) "Person Transfer GAN to Bridge Domain Gap for Person Re-Identification" [paper]

  • Rank-1 (Market1501) : 38.6(source domain : DukeMTMC)


[CVPR2017]

1) "Person Re-Identification in the Wild" [paper]


:high_brightness: [Market-1501 Leaderboard]

:high_brightness: [Collection]



ETC

Gait-based approach

These are papers that search for "gait" and "person re-identification" in google scholar.

1) "Person Re-identification by Video Ranking", ECCV2014 [paper]

2) "Person Re-identification using View-dependent Score-level Fusion of Gait and Color Features", ICPR 2012 [paper]

3) "Enhancing Person Re-identification by Integrating Gait Biometric", Neurocomputing 2015 [paper]

4) "A Hierarchical Method Combining Gait and Phase of Motion with Spatiotemporal Model for Person Re-identification", Pattern Recognition Letters 2012 [paper]

5) "Person Re-Identification by Discriminative Selection in Video Ranking", TPAMI 2016 [paper]

6) "A Spatio-temporal Appearance Representation for Video-based Pedestrian Re-identification", ICCV 2015 [paper]

7) "Gait-Assisted Person Re-identification in Wide Area Surveillance", ACCV 2014 [paper]

8) "Person Re-identification by Unsupervised Video Matching", Pattern Recognition 2017 [paper]

9) "Swiss-System Based Cascade Ranking for Gait-Based Person Re-identification", AAAI 2015 [paper]

10) "Person Re-identification using Height-based Gait in Colour Depth Camera", ICIP 2013 [paper]

11) "Review of person re-identification techniques", IET Computer vision 2013 [paper]

12) "Person Re-identification in Appearance Impaired Scenarios", arXiv 2016 [paper]

13) "Learning Compact Appearance Representation for Video-based Person Re-Identification", arXiv 2017 [paper]

14) "Recurrent Attention Models for Depth-Based Person Identification", CVPR 2016 [paper]

15) "Towards view-point invariant Person Re-identification via fusion of Anthropometric and Gait Features from Kinect measurements", VISAPP 2017 [paper]

16) "Person Re-identification in Frontal Gait Sequences via Histogram of Optic Flow Energy Image", ICACIVS 2016 [paper]

17) "Learning Bidirectional Temporal Cues for Video-based Person Re-Identification", TCSVT 2016 [paper]



Reference

https://github.com/handong1587/handong1587.github.io/blob/master/_posts/deep_learning/2015-10-09-re-id.md


推荐阅读