AlignedReID: Surpassing Human-Level Performance in Person Re-Identification
arXiv:1711.08184
Abstract
In this paper, we propose a novel method called AlignedReID that extracts a global feature which is jointly learned with local features. Global feature learning benefits greatly from local feature learning, which performs an alignment/matching by calculating the shortest path between two sets of local features, without requiring extra supervision. After the joint learning, we only keep the global feature to compute the similarities between images. Our method achieves rank-1 accuracy of 94.4% on Market1501 and 97.8% on CUHK03, outperforming state-of-the-art methods by a large margin. We also evaluate human-level performance and demonstrate that our method is the first to surpass human-level performance on Market1501 and CUHK03, two widely used Person ReID datasets.
9 pages, 8 figures
References in corpus (12)
- Person Re-identification: Past, Present and Future
- GLAD: Global-Local-Alignment Descriptor for Pedestrian Retrieval
- Deep Transfer Learning for Person Re-identification
- Pose Invariant Embedding for Deep Person Re-identification
- Margin Sample Mining Loss: A Deep Learning Based Method for Person Re-identification
- HydraPlus-Net: Attentive Deep Features for Pedestrian Analysis
- Multi-Target Tracking in Multiple Non-Overlapping Cameras using Constrained Dominant Sets
- Deep Mutual Learning
- Scalable Person Re-identification on Supervised Smoothed Manifold
- Deep Ranking Model by Large Adaptive Margin Learning for Person Re-identification
- Neural Person Search Machines
- Cross Domain Knowledge Transfer for Person Re-identification
Cited by in corpus (32)
- Bag of Tricks and A Strong Baseline for Deep Person Re-identification
- Unsupervised Embedding Learning via Invariant and Spreading Instance Feature
- Group Re-Identification with Multi-grained Matching and Integration
- Domain Adaptive Person Re-Identification via Coupling Optimization
- Identity-Guided Human Semantic Parsing for Person Re-Identification
- EANet: Enhancing Alignment for Cross-Domain Person Re-identification
- Dynamic Dual-Attentive Aggregation Learning for Visible-Infrared Person Re-Identification
- Beyond Triplet Loss: Person Re-identification with Fine-grained Difference-aware Pairwise Loss
- Generalizable Person Re-identification with Relevance-aware Mixture of Experts
- Intra-Inter Camera Similarity for Unsupervised Person Re-Identification
- Hetero-Center Loss for Cross-Modality Person Re-Identification
- FlowMOT: 3D Multi-Object Tracking by Scene Flow Association
- Attention: A Big Surprise for Cross-Domain Person Re-Identification
- Re-identification = Retrieval + Verification: Back to Essence and Forward with a New Metric
- Black Re-ID: A Head-shoulder Descriptor for the Challenging Problem of Person Re-Identification
- A Person Re-identification Data Augmentation Method with Adversarial Defense Effect
- ArTIST: Autoregressive Trajectory Inpainting and Scoring for Tracking
- Multi-Scale Body-Part Mask Guided Attention for Person Re-identification
- Apparel-invariant Feature Learning for Apparel-changed Person Re-identification
- Unsupervised Disentanglement GAN for Domain Adaptive Person Re-Identification
- Re-Identification Supervised Texture Generation
- Combining Two Adversarial Attacks Against Person Re-Identification Systems
- Unified Batch All Triplet Loss for Visible-Infrared Person Re-identification
- Improving Skeleton-based Action Recognitionwith Robust Spatial and Temporal Features
- Perceive Where to Focus: Learning Visibility-aware Part-level Features for Partial Person Re-identification
- Domain Adaptive Person Re-Identification via Camera Style Generation and Label Propagation
- Single Camera Training for Person Re-identification
- VMRFANet:View-Specific Multi-Receptive Field Attention Network for Person Re-identification
- Learning Incremental Triplet Margin for Person Re-identification
- Bridging the Distribution Gap of Visible-Infrared Person Re-identification with Modality Batch Normalization
- A Generic Object Re-identification System for Short Videos
- Pose Invariant Person Re-Identification using Robust Pose-transformation GAN