Research map / Label-efficient and robust learning / Self-supervised and contrastive
Silhouette-based Gait Foundation Model
Dingqiang Ye, Chao Fan, Kartik Narayan, Bingzhe Wu, Chengwen Luo, Jianqiang Li, Vishal Patel
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 |
|---|---|---|---|
| use part-aware student instead of cropping (ablation) | FoundationGait-0.13B · Gait3D self-supervised | rank-1: 17.2 -> 48.0 | Improved |
| use part-aware student instead of masking (ablation) | FoundationGait-0.13B · Gait3D self-supervised | rank-1: 38.1 -> 48.0 | Improved |
| expand P from [1] to [1,2,4,8] (ablation) | FoundationGait-0.13B · Gait3D self-supervised | rank-1: 39.6 -> 48.0 | Improved |
| use P=[1,4] instead of [1] (partial expansion) (ablation) | FoundationGait-0.13B · Gait3D self-supervised | rank-1: 39.6 -> 45.3 | Improved |
| add part-aware training to supervised DeepGaitV2-0.13B (ablation) | DeepGaitV2-0.13B · Gait3D | rank-1: 70.8 -> 76.3 | Improved |
| add part-aware training to supervised DeepGaitV2-0.13B (ablation) | DeepGaitV2-0.13B · CASIA-B NM | rank-1: 79.2 -> 90.0 | Improved |
| add part-aware training to DeepGaitV2-0.03B (ablation) | DeepGaitV2-0.03B · Gait3D | rank-1: 73.3 -> 74.3 | Improved |
| scale GaitSSB from 0.006B to 0.13B on GaitLU-1M | GaitSSB · OU-MVLP self-supervised | rank-1: 37.2 -> 26.3 | Hurt |
| scale GaitSSB from 0.006B to 0.13B on WebGait-2M | GaitSSB · OU-MVLP self-supervised | rank-1: 44.4 -> 38.9 | Hurt |
| scale FoundationGait from 0.03B to 0.13B | FoundationGait · Gait3D self-supervised | rank-1: 41.1 -> 48.0 | Improved |
| pretrain on diverse WebGait-1M instead of GaitLU-1M at equal scale (ablation) | FoundationGait-0.13B · OU-MVLP self-supervised | rank-1: 35.5 -> 58.6 | Improved |
| pretrain on WebGait-1M instead of GaitLU-1M at equal scale (ablation) | GaitSSB-0.13B · OU-MVLP self-supervised | rank-1: 26.3 -> 34.5 | Improved |
Closest papers by meaning
- BigGait: Learning Gait Representation You Want by Large Vision Models (CVPR 2024 · similarity 0.89)
- QAGait: Revisit Gait Recognition from a Quality Perspective (AAAI 2024 · similarity 0.87)
- BiggerGait: Unlocking Gait Recognition with Layer-wise Representations from Large Vision Models (NeurIPS 2025 · similarity 0.86)
- Cross-Covariate Gait Recognition: A Benchmark (AAAI 2024 · similarity 0.86)
- Vocabulary-Guided Gait Recognition (NeurIPS 2025 · similarity 0.86)
- SkeletonGait: Gait Recognition Using Skeleton Maps (AAAI 2024 · similarity 0.86)
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.
