Research map / Architectures and efficiency / CNN design
Robust Facial Landmark Detection via a Fully-Convolutional Local-Global Context Network
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 |
|---|---|---|---|
| add PCA-based 2D shape-model fitting to the network heatmaps (ablation) | Local-global context network · iBUG | median IOD-normalized MAE: 5.55 -> 5.29 / 4.36 -> 4.18 | Helped |
| add PCA-based 2D shape-model fitting to the network heatmaps (ablation) | Local-global context network · LFPW + HELEN | median IOD-normalized MAE: 3.04 -> 2.86 / 2.34 -> 2.21 | Helped |
| add PCA-based 2D shape-model fitting to the network heatmaps (ablation) | Local-global context network · Menpo frontal | median image-size-normalized MAE: 2.27 -> 2.14 / 1.90 -> 1.79 | Helped |
| replace square-error heatmap loss with pixel-wise cross-entropy loss (ablation) | Local-global context network · 300-W | landmark detection performance | Hurt |
| replace square-error heatmap loss with pixel-wise cross-entropy loss (ablation) | Local-global context network · Kaggle facial keypoint challenge | landmark detection performance | Mixed |
| use RGB or HSV input instead of grayscale (ablation) | Local-global context network · Facial landmark detection experiments | landmark detection performance | No clear effect |
| replace dilated convolutions with deformable convolutions (ablation) | Global-context subnet · Facial landmark detection experiments | performance | No clear effect |
Closest papers by meaning
- Attention-Driven Cropping for Very High Resolution Facial Landmark Detection (CVPR 2020 · similarity 0.89)
- Deep Structured Prediction for Facial Landmark Detection (NeurIPS 2019 · similarity 0.89)
- Semantic Alignment: Finding Semantically Consistent Ground-Truth for Facial Landmark Detection (CVPR 2019 · similarity 0.88)
- A Deep Regression Architecture With Two-Stage Re-Initialization for High Performance Facial Landmark Detection (CVPR 2017 · similarity 0.88)
- Robust Facial Landmark Detection via Occlusion-Adaptive Deep Networks (CVPR 2019 · similarity 0.88)
- Domain Generalized Medical Landmark Detection via Robust Boundary-Aware Pre-Training (AAAI 2025 · similarity 0.85)
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.