Research map / Label-efficient and robust learning / Recognition and long-tail
OpenGait: Revisiting Gait Recognition Towards Better Practicality
Chao Fan, Junhao Liang, Chuanfu Shen, Saihui Hou, Yongzhen Huang, Shiqi Yu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Remove MGP and the multi-scale mechanism in HPP (ablation) | GaitSet · Gait3D | rank-1 accuracy (%): 44.3 -> 45.8 | Improved |
| Replace FConv with regular convolution (ablation) | GaitPart · Gait3D | rank-1 accuracy (%): 29.2 -> 36.2 (+7.0%) | Improved |
| Remove the local feature branch (ablation) | GaitGL · Gait3D | rank-1 accuracy (%): 31.4 -> 32.2 | Improved |
| Add the SMPL branch (ablation) | SMPLGait with ResNet-like silhouette backbone · Gait3D | rank-1 accuracy (%): 56.5 -> 55.2 | Hurt |
| Add BNNeck (ablation) | GaitBase · GREW | rank-1 accuracy (%): 57.9 -> 60.1 | Improved |
| Add data augmentation (ablation) | GaitBase · Gait3D | rank-1 accuracy (%): 54.7 -> 64.6 (+9.9%) | Improved |
| Add data augmentation (ablation) | GaitBase · OU-MVLP | rank-1 accuracy (%): 90.8 -> 90.0 (-0.8%) | Hurt |
| Replace ResNet9 with ResNet50 (ablation) | GaitBase · CASIA-B* | rank-1 accuracy (%): -5.2%, -6.8%, -9.6% on NM, BG and CL | Hurt |
| Use ResNet50 instead of ResNet9 (ablation) | GaitBase · Gait3D | rank-1 accuracy (%): +1.9% | No clear effect |
| Add horizontal flipping (ablation) | GaitBase · Gait3D | rank-1 accuracy (%): 54.7 -> 59.5 | Improved |
| Add rotation augmentation (ablation) | GaitBase · CASIA-B* | rank-1 accuracy (%): 86.8 -> 87.4 | Improved |
| Add perspective transformation (ablation) | GaitBase · Gait3D | rank-1 accuracy (%): 54.7 -> 58.9 | Improved |
Closest papers by meaning
- BigGait: Learning Gait Representation You Want by Large Vision Models (CVPR 2024 · similarity 0.90)
- BiggerGait: Unlocking Gait Recognition with Layer-wise Representations from Large Vision Models (NeurIPS 2025 · similarity 0.89)
- Cross-Covariate Gait Recognition: A Benchmark (AAAI 2024 · similarity 0.88)
- RA-GAR: A Richly Annotated Benchmark for Gait Attribute Recognition (AAAI 2025 · similarity 0.88)
- Vocabulary-Guided Gait Recognition (NeurIPS 2025 · similarity 0.88)
- SkeletonGait: Gait Recognition Using Skeleton Maps (AAAI 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
