FD-GAN: Pose-guided Feature Distilling GAN for Robust Person Re-identification

Abstract

Person re-identification (reID) is an important task that requires to retrieve a person's images from an image dataset, given one image of the person of interest. For learning robust person features, the pose variation of person images is one of the key challenges. Existing works targeting the problem either perform human alignment, or learn human-region-based representations. Extra pose information and computational cost is generally required for inference. To solve this issue, a Feature Distilling Generative Adversarial Network (FD-GAN) is proposed for learning identity-related and pose-unrelated representations. It is a novel framework based on a Siamese structure with multiple novel discriminators on human poses and identities. In addition to the discriminators, a novel same-pose loss is also integrated, which requires appearance of a same person's generated images to be similar. After learning pose-unrelated person features with pose guidance, no auxiliary pose information and additional computational cost is required during testing. Our proposed FD-GAN achieves state-of-the-art performance on three person reID datasets, which demonstrates that the effectiveness and robust feature distilling capability of the proposed FD-GAN.

arXiv commentAccepted in Proceedings of 32nd Conference on Neural Information Processing Systems (NeurIPS 2018). Code available: https://github.com/yxgeee/FD-GAN

Cited by in corpus (61)

• Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID
• MHSA-Net: Multi-Head Self-Attention Network for Occluded Person Re-Identification
• Omni-Scale Feature Learning for Person Re-Identification
• Learning Feature Recovery Transformer for Occluded Person Re-identification
• Incomplete Descriptor Mining with Elastic Loss for Person Re-Identification
• Joint Discriminative and Generative Learning for Person Re-identification
• Parameter-Efficient Person Re-identification in the 3D Space
• Learning Multi-Attention Context Graph for Group-Based Re-Identification
• Feature Completion Transformer for Occluded Person Re-identification
• Learning Generalisable Omni-Scale Representations for Person Re-Identification
• High-Order Information Matters: Learning Relation and Topology for Occluded Person Re-Identification
• Learning Invariance from Generated Variance for Unsupervised Person Re-identification
• Disentangled Representations for Short-Term and Long-Term Person Re-Identification
• Exploiting Robust Unsupervised Video Person Re-identification
• Learning Shape Representations for Clothing Variations in Person Re-Identification
• Diverse Part Discovery: Occluded Person Re-identification with Part-Aware Transformer
• A Shape Transformation-based Dataset Augmentation Framework for Pedestrian Detection
• Hi-CMD: Hierarchical Cross-Modality Disentanglement for Visible-Infrared Person Re-Identification
• Style Normalization and Restitution for Generalizable Person Re-identification
• Rotate-and-Render: Unsupervised Photorealistic Face Rotation from Single-View Images
• Joint Disentangling and Adaptation for Cross-Domain Person Re-Identification
• Optimizing speed/accuracy trade-off for person re-identification via knowledge distillation
• Cross-Resolution Adversarial Dual Network for Person Re-Identification and Beyond
• Self-distillation with Batch Knowledge Ensembling Improves ImageNet Classification
• Progressive Sample Mining and Representation Learning for One-Shot Person Re-identification with Adversarial Samples
• Recover and Identify: A Generative Dual Model for Cross-Resolution Person Re-Identification
• Cross-Spectrum Dual-Subspace Pairing for RGB-infrared Cross-Modality Person Re-Identification
• Vehicle Re-identification in Aerial Imagery: Dataset and Approach
• Self-supervising Fine-grained Region Similarities for Large-scale Image Localization
• Cloth-Changing Person Re-identification from A Single Image with Gait Prediction and Regularization
• Pose-guided Inter- and Intra-part Relational Transformer for Occluded Person Re-Identification
• Real-world Person Re-Identification via Degradation Invariance Learning
• Semantics-Aligned Representation Learning for Person Re-identification
• An Introduction to Person Re-identification with Generative Adversarial Networks
• Joint Generative and Contrastive Learning for Unsupervised Person Re-identification
• Adversarial Metric Attack and Defense for Person Re-identification
• DF^2AM: Dual-level Feature Fusion and Affinity Modeling for RGB-Infrared Cross-modality Person Re-identification
• Attention Deep Model with Multi-Scale Deep Supervision for Person Re-Identification
• Global Distance-distributions Separation for Unsupervised Person Re-identification
• Dual Gaussian-based Variational Subspace Disentanglement for Visible-Infrared Person Re-Identification
• Improved Mutual Mean-Teaching for Unsupervised Domain Adaptive Re-ID
• Conditional Adversarial Generative Flow for Controllable Image Synthesis
• Unity Style Transfer for Person Re-Identification
• PMC-GANs: Generating Multi-Scale High-Quality Pedestrian with Multimodal Cascaded GANs
• Multi-Object Tracking with Deep Learning Ensemble for Unmanned Aerial System Applications
• AttKGCN: Attribute Knowledge Graph Convolutional Network for Person Re-identification
• Discriminability Distillation in Group Representation Learning
• Video Reenactment as Inductive Bias for Content-Motion Disentanglement
• Unsupervised Disentanglement GAN for Domain Adaptive Person Re-Identification
• Relation Network for Person Re-identification
• Joint Discriminative and Metric Embedding Learning for Person Re-Identification
• Deep Multi-Index Hashing for Person Re-Identification
• Anchor-Free Person Search
• Deep High-Resolution Representation Learning for Cross-Resolution Person Re-identification
• Efficient Person Search: An Anchor-Free Approach
• Disentangling Latent Space for VAE by Label Relevant/Irrelevant Dimensions
• Feature Completion for Occluded Person Re-Identification
• Pose-Guided Feature Learning with Knowledge Distillation for Occluded Person Re-Identification
• When Relation Networks meet GANs: Relation GANs with Triplet Loss
• Hierarchical Modes Exploring in Generative Adversarial Networks
• Unsupervised Domain Adaptive Person Re-Identification via Human Learning Imitation