Research map / Detection and segmentation / CNN features
Pixel-in-Pixel Net: Towards Efficient Facial Landmark Detection in the Wild
Haibo Jin, Shengcai Liao, Ling Shao
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 |
|---|---|---|---|
| use network stride 32 instead of 16 (ablation) | PIPNet w/o NRM · WFLW validation | NME (%): 4.03 -> 3.94 | Improved |
| use network stride 128 instead of 32 (ablation) | PIPNet w/o NRM · WFLW validation | NME (%): 3.94 -> 4.10 | Hurt |
| add neighbor regression module (NRM) (ablation) | PIPNet · WFLW validation | NME (%) | Improved |
| replace standard self-training with self-training with curriculum (UDA) (ablation) | PIPNet · WFLW-68 | NME (%): 7.45 -> 7.28 | Improved |
| replace standard self-training with self-training with curriculum (UDA) (ablation) | PIPNet · COFW-68 | NME (%): 4.34 -> 4.28 | Improved |
| replace standard self-training with self-training with curriculum (GSSL, CelebA) (ablation) | PIPNet · 300W | NME (%): 3.27 -> 3.23 | Improved |
| add standard self-training with unlabeled CelebA (GSSL) | PIPNet · COFW-68 | NME (%): 4.55 -> 4.32 | Improved |
| add DANN domain adaptation (UDA) | PIPNet · 300W | NME (%): 3.36 -> 3.42 | Hurt |
| add DANN domain adaptation (UDA) | PIPNet · COFW-68 | NME (%): 4.55 -> 4.55 | No clear effect |
| use CelebA (GSSL) instead of target-domain unlabeled data (UDA) | PIPNet STC · WFLW-68 | NME (%): 7.28 -> 7.53 | Hurt |
| use ResNet-101 instead of ResNet-18 backbone | PIPNet · COFW | NME (%): 3.31 -> 3.08 | Improved |
| use ResNet-101 instead of ResNet-18 backbone | PIPNet · 300VW category 1 | NME (%): 3.04 -> 3.03 | No clear effect |
Closest papers by meaning
- 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)
- Facial Landmark Detection with Tweaked Convolutional Neural Networks (TPAMI 2017 · similarity 0.89)
- Attention-Driven Cropping for Very High Resolution Facial Landmark Detection (CVPR 2020 · similarity 0.88)
- Facial Landmark Detection: A Literature Survey (IJCV 2018 · similarity 0.88)
- POPoS: Improving Efficient and Robust Facial Landmark Detection with Parallel Optimal Position Search (AAAI 2025 · similarity 0.86)
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.
