Research map / Label-efficient and robust learning / Recognition and long-tail
Robustness Disparities in Face Detection
Samuel Dooley, George Z Wei, Tom Goldstein, John P Dickerson
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 |
|---|---|---|---|
| compare commercial against academic models | commercial face detectors · all four datasets | average precision under corruption | Improved |
| use MogFace instead of commercial models | MogFace · UTKFace | average precision under corruption | Improved |
| use MogFace instead of TinaFace and Yolo5Face | MogFace · all datasets except CCD | average precision under corruption | Improved |
| apply zoom blur corruption | all six models · CCD and MIAP | average precision | Hurt |
| apply elastic transform corruption | four of six models · UTKFace | average precision | Hurt |
| apply noise corruptions | TinaFace and Yolo5Face · UTKFace | average precision | Hurt |
| perceived feminine vs masculine presentation | six face detectors · 24 dataset-model combinations | noise-induced AP change: 19 of 24 biased against masculine; 4 against feminine; 1 none | Mixed |
| darker vs lighter skin type | all models · CCD | noise robustness AP | Hurt |
| dim vs bright lighting | all models except YOLO5Face · CCD | noise robustness AP | Hurt |
| dim lighting on skin type gap | six face detectors · CCD | skin-type AP gap | Improved |
| older vs middle-aged subjects | six face detectors · Adience, MIAP, UTKFace | noise robustness AP | Hurt |
| commercial vs academic models in disparity magnitude | commercial face detectors · all datasets | demographic disparity | No clear effect |
Closest papers by meaning
- Longitudinal Study of Automatic Face Recognition (TPAMI 2017 · similarity 0.88)
- FACET: Fairness in Computer Vision Evaluation Benchmark (ICCV 2023 · similarity 0.87)
- Benchmarking Algorithmic Bias in Face Recognition: An Experimental Approach Using Synthetic Faces and Human Evaluation (ICCV 2023 · similarity 0.87)
- Towards causal benchmarking of bias in face analysis algorithms (ECCV 2020 · similarity 0.86)
- A Call to Reflect on Evaluation Practices for Age Estimation: Comparative Analysis of the State-of-the-Art and a Unified Benchmark (CVPR 2024 · similarity 0.84)
- FAIR-FER: A Latent Alignment Approach for Mitigating Bias in Facial Expression Recognition (Student Abstract) (AAAI 2024 · similarity 0.83)
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.
