Research map / Label-efficient and robust learning / Domain adaptation and generalization
Fine-grained Unsupervised Domain Adaptation for Gait Recognition
Kang Ma, Ying Fu, Dezhi Zheng, Yunjie Peng, Chunshui Cao, Yongzhen Huang
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 |
|---|---|---|---|
| add Cycle Temporal Shift convolution instead of 2D convolution (ablation) | UDA(STANet) · CASIA-B CL (OUMVLP->CASIA-B) | Rank-1: 62.5 -> 68.2 | Improved |
| add Cycle Temporal Shift convolution instead of 2D convolution (ablation) | UDA(STANet) · CASIA-B NM (OUMVLP->CASIA-B) | Rank-1: 91.7 -> 90.6 | Hurt |
| add global context block instead of temporal max pooling (ablation) | UDA(STANet) · CASIA-B (OUMVLP->CASIA-B) | Rank-1 mean: 76.1 -> 80.6 | Improved |
| add part confidence scores for pseudo-label generation (ablation) | UDA(STANet) · CASIA-B (OUMVLP->CASIA-B) | Rank-1 mean: 78.9 -> 80.6 | Improved |
| increase number of horizontal parts from 2 to 32 (ablation) | UDA(STANet) · CASIA-B (OUMVLP->CASIA-B) | Rank-1 mean: 61.1 -> 80.6 | Improved |
| UDA framework versus fine-tuning UDA method TranND | CASIA-B CL · CASIA-B CL (OUMVLP->CASIA-B) | Rank-1 mean: 15.8 -> 68.2 | Improved |
| swap backbone from STANet to GaitSet in UDA framework | UDA framework · CASIA-B (OUMVLP->CASIA-B) | Rank-1 mean: NM 90.6 -> 93.2; CL 68.2 -> 36.7 | Mixed |
| STANet backbone versus GaitSet and GaitGL in UDA, CASIA-B to OUMVLP | UDA framework · OUMVLP (CASIA-B->OUMVLP) | Rank-1 mean: 69.5 -> 82.9 vs GaitSet | Improved |
| UDA transfer from OUMVLP to GREW versus supervised training | UDA(STANet) · GREW | Rank-1: 41.3 -> 26.2 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- GaitSet — GaitSet: Regarding Gait as a Set for Cross-View Gait Recognition (AAAI 2019) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Learning Gait Representation From Massive Unlabelled Walking Videos: A Benchmark (TPAMI 2023 · similarity 0.90)
- Gait Recognition in the Wild: A Large-Scale Benchmark and NAS-Based Baseline (TPAMI 2025 · similarity 0.89)
- A Comprehensive Study on Cross-View Gait Based Human Identification with Deep CNNs (TPAMI 2016 · similarity 0.88)
- Cross-Covariate Gait Recognition: A Benchmark (AAAI 2024 · similarity 0.88)
- Learning Visual Prompt for Gait Recognition (CVPR 2024 · similarity 0.87)
- BiggerGait: Unlocking Gait Recognition with Layer-wise Representations from Large Vision Models (NeurIPS 2025 · 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.
