Research map / 3D vision / Face and shape
3D morphable face model regression
Regresses morphable face parameters from images using deep networks, graph convolutions and implicit models.
80 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Face and shape, keyword cluster “3d face · face reconstruction · faces”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
19 papers in 2023–24 and 5 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 5.2% to 1.4%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Face Alignment Across Large Poses: A 3D Solution (CVPR 2016 · 1,211 citations)
- 2017 · Regressing Robust and Discriminative 3D Morphable Models With a Very Deep Neural Network (CVPR 2017 · 534 citations)
- 2018 · Nonlinear 3D Face Morphable Model (CVPR 2018 · 423 citations)
- 2019 · Learning to Regress 3D Face Shape and Expression From an Image Without 3D Supervision (CVPR 2019 · 346 citations)
- 2020 · FaceScape: A Large-Scale High Quality 3D Face Dataset and Detailed Riggable 3D Face Prediction (CVPR 2020 · 319 citations)
- 2021 · i3DMM: Deep Implicit 3D Morphable Model of Human Heads (CVPR 2021 · 122 citations)
- 2022 · FaceVerse: A Fine-Grained and Detail-Controllable 3D Face Morphable Model From a Hybrid Dataset (CVPR 2022 · 101 citations)
- 2023 · A Hierarchical Representation Network for Accurate and Detailed Face Reconstruction From In-the-Wild Images (CVPR 2023 · 88 citations)
- 2024 · 3D Face Reconstruction with the Geometric Guidance of Facial Part Segmentation (CVPR 2024 · 37 citations)
- 2025 · Learning Person-Specific Animatable Face Models from In-the-Wild Images via a Shared Base Model (CVPR 2025 · 0 citations)
- 2026 · NICE: Neural Implicit Craniofacial Model for Orthognathic Surgery Prediction (AAAI 2026 · 0 citations)
Earliest papers here
- Face Alignment Across Large Poses: A 3D Solution (CVPR 2016 · 1,211 citations)
- A 3D Morphable Model Learnt From 10,000 Faces (CVPR 2016 · 376 citations)
- Large-Pose Face Alignment via CNN-Based Dense 3D Model Fitting (CVPR 2016 · 302 citations)
What has been tried
77 papers here report 674 measured changes (a component added, removed or swapped, compared with a baseline): helped 77%, no clear effect 5%, hurt 16%, mixed 2%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| add PRDL | Part IoU benchmark · Part IoU average (%) | 72.08 -> 73.82 | Helped | 3D Face Reconstruction with the Geometric Guidance of Facial Part Segmentation (CVPR 2024 · 37 citations) |
| replace sphere tracing with NeuS-style volume rendering | 20 Kinect sequences · L1-Ch. | 0.0033 -> 0.0024 | Helped | MonoNPHM: Dynamic Head Reconstruction from Monocular Videos (CVPR 2024 · 24 citations) |
| remove StyleGAN2 coefficient refinement | Wild · RMSE | 0.0618 -> 0.0517 | Hurt | Makeup Prior Models for 3D Facial Makeup Estimation and Applications (CVPR 2024 · 9 citations) |
| Add the style loss | Overall-Average · CLIP average | 0.9470 -> 0.9491 | Helped | LeGO: Leveraging a Surface Deformation Network for Animatable Stylized Face Generation with One Example (CVPR 2024 · 8 citations) |
| Apply expression coefficients globally instead of locally | Novel 3D performances · average fitting error (mm) | 0.31 -> 0.86 | Hurt | Anatomically Constrained Implicit Face Models (CVPR 2024 · 7 citations) |
| add the multi-view gaze-representation swapping constraint to the dual-encoder training | ETH-XGaze · gaze estimation error (°), 50/100/200-shot average | 20.34 -> 8.17 | Helped | Unsupervised Gaze Representation Learning from Multi-view Face Images (CVPR 2024 · 7 citations) |
| replace FaRL ViT with Resnet50 | DAD-3DHeads · NMLC (full) | 1.91 -> 2.23 | Hurt | FaceLift: Semi-supervised 3D Facial Landmark Localization (CVPR 2024 · 4 citations) |
| Add tri-plane warping as data augmentation when training the encoder | facial point manipulation · manipulation perception | Helped | 3D-Aware Face Editing via Warping-Guided Latent Direction Learning (CVPR 2024 · 3 citations) | |
| add cycle-consistent shape loss Lcc (CSG) | MPIE-LP self-reenactment · FID | 23.9 -> 18.1 | Helped | Pose Adapted Shape Learning for Large-Pose Face Reenactment (CVPR 2024 · 3 citations) |
| add silhouette vertex re-projection loss | 300-W · boundary RMSE error | 0.0616 -> 0.0559 | Helped | Learning Person-Specific Animatable Face Models from In-the-Wild Images via a Shared Base Model (CVPR 2025 · 0 citations) |
| Increase the number of keypoints from 4 to 18 | 3DFAW, 1 view · Chamfer distance (mm) | 1.92 -> 1.25 | Helped | GLVD: Guided Learned Vertex Descent (NeurIPS 2025 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 13 papers of this approach from 2024–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- FFHQ · 3 of 13 papers
- Dad-3dheads · 2 of 13 papers
- 300W · 2 of 13 papers
- MEAD · 2 of 13 papers
- CelebV-HQ · 2 of 13 papers
- VoxCeleb1 · 2 of 13 papers
Baselines
- DECA · 5 of 13 papers
- 3DDFA-v2 · 3 of 13 papers
- HRN · 2 of 13 papers
- ArcFace · 2 of 13 papers
- AdaFace · 2 of 13 papers
- SynergyNet · 2 of 13 papers
- FaceScape · 2 of 13 papers
- FaceVerse · 2 of 13 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 24 papers; from OpenAlex and the papers' affiliations).
- Studios · 2 papers
- Disney Research · 2 papers
- Chinese Academy of Sciences · 2 papers
- University of Chinese Academy of Sciences · 2 papers
- Synthesia · 2 papers
- Technical University of Munich · 2 papers
- University College London · 2 papers
- Shanghai Jiao Tong University · 2 papers
- Beihang University · 2 papers
Use it
Download BibTeX of 35 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Face reflectance and relightable rendering (85 papers)
- Speech-driven 3D facial animation (32 papers)
Elsewhere in the topic
- Neural implicit surface reconstruction (131 papers)
- Gaussian and implicit head avatars (115 papers)
- Textured mesh generation (111 papers)
- Primitive-based shape abstraction (99 papers)
- Single-view 3D reconstruction (96 papers)
- Full-body avatar reconstruction (78 papers)
- Articulated and non-rigid shape learning (75 papers)
- 3D shape generative modeling (74 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).