Research map / Detection and segmentation / CNN features
Deep Structured Prediction for Facial Landmark Detection
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 |
|---|---|---|---|
| use the CNN-CRF structured prediction method instead of LAB | 4-stack facial landmark detection · 300W Full (inter-ocular distance) | NME (%): 3.49 -> 3.30 | Helped |
| use CNN-CRF instead of SAN | facial landmark detection · 300W-test | NME: 2.86 -> 2.21 | Helped |
| use CNN-CRF instead of CE-CLM | facial landmark detection · Menpo-frontal | NME: 2.78 -> 2.01 | Helped |
| use CNN-CRF instead of CE-CLM | facial landmark detection · Menpo-profile | NME: 4.63 -> 3.03 | Helped |
| use CNN-CRF instead of CE-CLM | facial landmark detection · COFW-68 test | NME: 3.36 -> 2.55 | Helped |
| use CNN-CRF instead of CFSS | facial landmark detection · 300VW-category1 | NME: 2.44 -> 1.91 | Helped |
| use CNN-CRF instead of CFSS | facial landmark detection · 300VW-category2 | NME: 2.49 -> 1.97 | Helped |
| use CNN-CRF instead of CFSS | facial landmark detection · 300VW-category3 | NME: 3.26 -> 2.50 | Helped |
| add L1 mean loss and Gaussian NLL loss to softmax cross entropy (ablation) | plain CNN · 300W-test | NME: 2.38 -> 2.30 | Helped |
| add L1 mean loss and Gaussian NLL loss to softmax cross entropy (ablation) | plain CNN · 300W-test | AUC: 65.9 -> 67.4 | Helped |
| train CNN and CRF jointly rather than separately (ablation) | CNN-CRF with proposed loss · 300W-test | NME: 2.23 -> 2.21 | Helped |
| train CNN and CRF jointly rather than separately (ablation) | CNN-CRF with proposed loss · 300W-test | AUC: 67.8 -> 68.1 | Helped |
Closest papers by meaning
- Continuous Landmark Detection With 3D Queries (CVPR 2023 · similarity 0.89)
- Learning to Detect 3D Facial Landmarks via Heatmap Regression with Graph Convolutional Network (AAAI 2022 · similarity 0.89)
- Robust Facial Landmark Detection via a Fully-Convolutional Local-Global Context Network (CVPR 2018 · similarity 0.89)
- Semantic Alignment: Finding Semantically Consistent Ground-Truth for Facial Landmark Detection (CVPR 2019 · similarity 0.88)
- Domain Generalized Medical Landmark Detection via Robust Boundary-Aware Pre-Training (AAAI 2025 · similarity 0.84)
- FaceLift: Semi-supervised 3D Facial Landmark Localization (CVPR 2024 · 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.