Research map / 3D vision / Face and shape
Gait Recognition via Disentangled Representation Learning
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 GaitNet instead of L-CRF | CASIA-B · NM cross-view average | Rank-1 accuracy: 67.8 -> 81.8 | Helped |
| Use GaitNet instead of LB | CASIA-B · Cross-view NM | Mean accuracy: 89.6 -> 93.9 | Helped |
| Use GaitNet instead of LB-subGEI | CASIA-B · Cross-view BG | Mean accuracy: 70.3 -> 82.6 | Helped |
| Use GaitNet instead of LB-subGEI | CASIA-B · Cross-view CL | Mean accuracy: 53.1 -> 63.2 | Helped |
| Use GaitNet instead of LB network | USF · Probe A | Identification accuracy: 96.7\pm0.5\% -> 99.5\pm 0.2\% | Helped |
| Use GaitNet instead of PE-LSTM | FVG · All protocol, TDR@1% FAR | TDR: 65.4 -> 81.2 | Helped |
| Use GaitNet instead of PE-LSTM | FVG · All protocol, TDR@5% FAR | TDR: 74.1 -> 87.8 | Helped |
| Add cross reconstruction loss (ablation) | GaitNet with conventional identity loss · CASIA-B, NM/CL cross-condition | Rank-1 accuracy: 56.0 -> 60.2 | Helped |
| Add gait similarity loss alongside cross reconstruction (ablation) | GaitNet with conventional identity loss · CASIA-B, NM/CL cross-condition | Rank-1 accuracy: 60.2 -> 85.6 | Helped |
| Use incremental identity loss rather than average-output identity loss (ablation) | GaitNet with both disentanglement losses · CASIA-B, NM/CL cross-condition | Rank-1 accuracy: 62.6 -> 85.6 | Helped |
| Use incremental identity loss rather than final-step identity loss (ablation) | GaitNet with both disentanglement losses · CASIA-B, NM/CL cross-condition | Rank-1 accuracy: 26.0 -> 85.6 | Helped |
| Use incremental identity loss rather than autoencoder identity loss (ablation) | GaitNet with both disentanglement losses · CASIA-B, NM/CL cross-condition | Rank-1 accuracy: 71.2 -> 85.6 | Helped |
Closest papers by meaning
- Exploring More from Multiple Gait Modalities for Human Identification (AAAI 2025 · similarity 0.88)
- On Denoising Walking Videos for Gait Recognition (CVPR 2025 · 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 0.87)
- Vocabulary-Guided Gait Recognition (NeurIPS 2025 · similarity 0.86)
- Cross-Covariate Gait Recognition: A Benchmark (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.