Research map / Image generation / Autoregressive and latent generation
Attention-Driven Cropping for Very High Resolution Facial Landmark Detection
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 attention-driven regional cropping and refinement | Facial landmark detection on 300-W · 300-W Common | NME: 2.91 -> 2.83 | Helped |
| Evaluate only the 51 high-resolution landmarks (ablation) | Attention-driven detector · 300-W Challenging | NME: 7.04 -> 5.68 | Helped |
| Evaluate only the 51 high-resolution landmarks (ablation) | Attention-driven detector · 300-W Full Set | NME: 4.23 -> 3.50 | Helped |
| Use attention-driven regional cropping and refinement | Facial landmark detection on 300-VW · 300-VW Category 1 | NME: 4.40 -> 4.17 | Helped |
| Evaluate only the 51 high-resolution landmarks (ablation) | Attention-driven detector · 300-VW Category 1 | NME: 4.17 -> 3.66 | Helped |
| Use attention-driven regional cropping and refinement | Facial landmark detection on 300-VW · 300-VW Category 2 | NME: 4.16 -> 3.89 | Helped |
| Evaluate only the 51 high-resolution landmarks (ablation) | Attention-driven detector · 300-VW Category 2 | NME: 3.89 -> 3.35 | Helped |
| Use attention-driven regional cropping and refinement | Facial landmark detection on 300-VW · 300-VW Category 3 | NME: 1.13 -> 7.28 | Hurt |
| Use attention-driven regional cropping and refinement | Landmark detector on the high-resolution test set · 256x256 | NME: 2.34 -> 2.26 | Helped |
| Use attention-driven regional cropping and refinement | Landmark detector on the high-resolution test set · 512x512 | NME: 2.34 -> 1.95 | Helped |
| Use attention-driven regional cropping and refinement | Landmark detector on the high-resolution test set · 2048x2048 | NME: 2.34 -> 1.94 | Helped |
| Use depthwise separable convolutions (ablation) | Attention-driven landmark detector · High-resolution test set 4096x4096 | NME: 2.38 -> 1.95 | Helped |
Closest papers by meaning
- Robust Facial Landmark Detection via a Fully-Convolutional Local-Global Context Network (CVPR 2018 · similarity 0.89)
- Joint Super-Resolution and Alignment of Tiny Faces (AAAI 2020 · similarity 0.87)
- Deep Structured Prediction for Facial Landmark Detection (NeurIPS 2019 · similarity 0.86)
- A Deep Regression Architecture With Two-Stage Re-Initialization for High Performance Facial Landmark Detection (CVPR 2017 · similarity 0.86)
- Win-Win: Training High-Resolution Vision Transformers from Two Windows (ICLR 2024 · similarity 0.85)
- POPoS: Improving Efficient and Robust Facial Landmark Detection with Parallel Optimal Position Search (AAAI 2025 · similarity 0.84)
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.