Research map / Label-efficient and robust learning / Recognition and long-tail
Open-Set Biometrics: Beyond Good Closed-Set Models
Yiyang Su, Minchul Kim, Feng Liu, Anil Jain, Xiaoming Liu
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 the open-set identification-detection loss and relative threshold minimization to AdaFace | AdaFace · IJB-S Surveillance-Booking | FNIR@0.1%FPIR: 45.98 -> 43.48 | Improved |
| add the proposed losses to SwinGait | SwinGait-3D · Gait3D | FNIR@1%FPIR: 83.50 -> 79.30 | Improved |
| add the proposed losses to SwinGait | SwinGait-3D · Gait3D | rank-1 accuracy: 72.4 -> 75.3 | Improved |
| add the proposed losses to CAL | CAL · MEVID | rank-1 accuracy: 52.5 -> 55.1 | Improved |
| add the proposed losses to GaitBase | GaitBase · GREW | rank-1 accuracy: 60.1 -> 60.1 | No clear effect |
| add RTM only (ablation) | GaitBase · Gait3D validation | FNIR@1%FPIR: 86.43 -> 85.81 | Improved |
| add Det loss on top of RTM (ablation) | GaitBase · Gait3D validation | FNIR@1%FPIR: 85.81 -> 85.16 | Improved |
| add Id loss on top of Det and RTM (ablation) | GaitBase · Gait3D validation | FNIR@1%FPIR: 85.16 -> 82.99 | Improved |
| increase RTM weight lambda from 4.0 to 8.0 (ablation) | GaitBase · Gait3D validation | rank-1 accuracy: 64.8 -> 61.6 | Hurt |
| use proposed loss instead of EVM score calibration on SwinGait | SwinGait-3D · Gait3D | FNIR@1%FPIR: 81.08 (EVM) -> 79.30 (Ours) | Improved |
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.
- AdaFace — AdaFace: Quality Adaptive Margin for Face Recognition (CVPR 2022) · better on 3 of 3 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
- Open-Set Recognition: A Good Closed-Set Classifier is All You Need (ICLR 2022 · similarity 0.85)
- The Devil is in the Wrongly-classified Samples: Towards Unified Open-set Recognition (ICLR 2023 · similarity 0.85)
- Open-Set Facial Expression Recognition (AAAI 2024 · similarity 0.84)
- All Beings Are Equal in Open Set Recognition (AAAI 2024 · similarity 0.83)
- GHOST: Gaussian Hypothesis Open-Set Technique (AAAI 2025 · similarity 0.82)
- Flexible Biometrics Recognition: Bridging the Multimodality Gap through Attention Alignment and Prompt Tuning (CVPR 2024 · similarity 0.82)
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.
