Research map / Label-efficient and robust learning / Recognition and long-tail
AnchorFace: An Anchor-based Facial Landmark Detector Across Large Poses
Read the paperCheck 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 |
|---|---|---|---|
| Replace direct coordinate regression with anchor-based regression and aggregation (ablation) | ShuffleNet-V2 landmark detector · AFLW Full | NME: 1.67 -> 1.56 | Helped |
| Use KMeans-generated anchor templates instead of hand-designed templates (ablation) | AnchorFace · AFLW Full | NME: 1.58 -> 1.56 | Helped |
| Use 24 rather than 3 anchor templates (ablation) | AnchorFace · AFLW Full | NME: 1.60 -> 1.56 | Helped |
| Use 48 rather than 24 anchor templates (ablation) | AnchorFace · AFLW Full | NME: 1.56 -> 1.58 | Hurt |
| Set anchor area to 56×56 instead of 112×112 (ablation) | AnchorFace · AFLW Full | NME: 1.58 -> 1.56 | Helped |
| Set anchor area to 56×56 instead of 28×28 (ablation) | AnchorFace · AFLW Full | NME: 1.58 -> 1.56 | Helped |
| Set anchor area to 56×56 instead of 14×14 (ablation) | AnchorFace · AFLW Full | NME: 1.58 -> 1.56 | Helped |
| Use a 7×7 anchor grid rather than a 3×3 grid (ablation) | AnchorFace · AFLW Full | NME: 1.58 -> 1.56 | Helped |
| Use a 7×7 anchor grid rather than a 5×5 grid (ablation) | AnchorFace · AFLW Full | NME: 1.57 -> 1.56 | Helped |
| Use a 13×13 rather than 7×7 anchor grid (ablation) | AnchorFace · AFLW Full | NME: 1.56 -> 1.56 | No clear effect |
| Use confidence-weighted voting rather than argmax selection (ablation) | AnchorFace · AFLW Full | NME: 1.58 -> 1.56 | Helped |
| Weight aggregated predictions by confidence instead of taking their unweighted mean (ablation) | AnchorFace · AFLW Full | NME: 1.61 -> 1.56 | Helped |
Closest papers by meaning
- HAMBox: Delving Into Mining High-Quality Anchors on Face Detection (CVPR 2020 · similarity 0.89)
- Improving Landmark Localization With Semi-Supervised Learning (CVPR 2018 · similarity 0.88)
- Joint Super-Resolution and Alignment of Tiny Faces (AAAI 2020 · similarity 0.88)
- LUVLi Face Alignment: Estimating Landmarks' Location, Uncertainty, and Visibility Likelihood (CVPR 2020 · similarity 0.88)
- FaceLift: Semi-supervised 3D Facial Landmark Localization (CVPR 2024 · similarity 0.87)
- POPoS: Improving Efficient and Robust Facial Landmark Detection with Parallel Optimal Position Search (AAAI 2025 · similarity 0.87)
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.