Research map / 3D vision / Face and shape
A Simple, Fast and Highly-Accurate Algorithm to Recover 3D Shape from 2D Landmarks on a Single Image
Ruiqi Zhao, Yan Wang, Aleix M. Martínez
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 a deep feed-forward network instead of linear shape-basis models for 2D-to-3D landmark lifting | 3D shape recovery from 2D landmarks · CMU Motion Capture (Subject 13) | reconstruction error: .0653 -> .0229 | Improved |
| use the deep network instead of Ramakrishna et al. on Subject 15 | 3D shape recovery from 2D landmarks · CMU Motion Capture (Subject 15) | reconstruction error: .0675 -> .0099 | Improved |
| use the deep network instead of Zhou et al. on BU-3DFE faces | 3D shape recovery from 2D landmarks · BU-3DFE Face | reconstruction error: 0.0053 -> 0.0040 | Improved |
| use the deep network instead of Zhou et al. on 3DFAW | 3D shape recovery from 2D landmarks · 3DFAW | reconstruction error: 0.0055 -> 0.0042 | Improved |
| use the deep network instead of Zhou et al. on cars | 3D shape recovery from 2D landmarks · FG3DCar | reconstruction error: 0.0042 -> 0.0022 | Improved |
| add the recurrent layer so one randomly missing landmark is estimated jointly with depth | 3D shape recovery from 2D landmarks · CMU Motion Capture (Subject 13) | reconstruction error: .0229 -> 0.0413 | Hurt |
| missing one landmark during training and testing on BU-3DFE faces | 3D shape recovery from 2D landmarks · BU-3DFE Face | reconstruction error: 0.0040 -> 0.006 | Hurt |
| add Gaussian noise (about 3% of object size) to input 2D landmarks | 3D shape recovery from 2D landmarks · CMU Motion Capture | reconstruction error | No clear effect |
Closest papers by meaning
- How Far Are We From Solving the 2D & 3D Face Alignment Problem? (And a Dataset of 230,000 3D Facial Landmarks) (ICCV 2017 · similarity 0.91)
- Large Pose 3D Face Reconstruction From a Single Image via Direct Volumetric CNN Regression (ICCV 2017 · similarity 0.91)
- Image-Based 3D Object Reconstruction: State-of-the-Art and Trends in the Deep Learning Era (TPAMI 2019 · similarity 0.91)
- Front2Back: Single View 3D Shape Reconstruction via Front to Back Prediction (CVPR 2020 · similarity 0.90)
- Learning Detailed Face Reconstruction From a Single Image (CVPR 2017 · similarity 0.90)
- Do 2D GANs Know 3D Shape? Unsupervised 3D Shape Reconstruction from 2D Image GANs (ICLR 2021 · similarity 0.89)
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.
