Research map / 3D vision / Face and shape
Single-view 3D reconstruction
Reconstructs 3D shape from one or few images with hierarchical priors and generalization to unseen classes.
96 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Face and shape, keyword cluster “3d shape · shapes · object”
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.
17 papers in 2023–24 and 21 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 4.7% to 5.7%.
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 · Structured Prediction of Unobserved Voxels From a Single Depth Image (CVPR 2016 · 199 citations)
- 2017 · Multi-View Supervision for Single-View Reconstruction via Differentiable Ray Consistency (CVPR 2017 · 571 citations)
- 2018 · Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling (CVPR 2018 · 479 citations)
- 2019 · Learning to Reconstruct People in Clothing From a Single RGB Camera (CVPR 2019 · 351 citations)
- 2020 · Implicit Functions in Feature Space for 3D Shape Reconstruction and Completion (CVPR 2020 · 532 citations)
- 2021 · Learning 3D Shape Feature for Texture-Insensitive Person Re-Identification (CVPR 2021 · 158 citations)
- 2022 · Neural Fields As Learnable Kernels for 3D Reconstruction (CVPR 2022 · 62 citations)
- 2023 · RealFusion: 360deg Reconstruction of Any Object From a Single Image (CVPR 2023 · 178 citations)
- 2024 · ZeroShape: Regression-based Zero-shot Shape Reconstruction (CVPR 2024 · 19 citations)
- 2025 · Robust 3D Shape Reconstruction in Zero-Shot from a Single Image in the Wild (CVPR 2025 · 2 citations)
- 2026 · SAM 3D: 3Dfy Anything in Images (CVPR 2026 · 1 citation)
Earliest papers here
- Structured Prediction of Unobserved Voxels From a Single Depth Image (CVPR 2016 · 199 citations)
- A Field Model for Repairing 3D Shapes (CVPR 2016 · 117 citations)
- Large-Scale Semantic 3D Reconstruction: An Adaptive Multi-Resolution Model for Multi-Class Volumetric Labeling (CVPR 2016 · 91 citations)
What has been tried
78 papers here report 799 measured changes (a component added, removed or swapped, compared with a baseline): helped 73%, no clear effect 7%, hurt 19%, 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 explicit geometric reasoning with depth and camera pretraining losses | OmniObject3D · FS@5 | 0.7797 -> 0.8169 | Helped | ZeroShape: Regression-based Zero-shot Shape Reconstruction (CVPR 2024 · 19 citations) |
| Integrate prior during early and late denoising stages | ShapeNet Chair · Chamfer Distance | 89.79 -> 79.26 | Helped | Bayesian Diffusion Models for 3D Shape Reconstruction (CVPR 2024 · 17 citations) |
| add per-view importance weighting | Keywest and School Bus (NAVI) · PSNR | 22.43 -> 25.87 | Helped | SHINOBI: Shape and Illumination using Neural Object Decomposition via BRDF Optimization In-the-wild (CVPR 2024 · 13 citations) |
| model unintended occluders | Stanford-ORB salt004 · albedo PSNR | 14.4 dB -> 19.3 dB | Helped | Eclipse: Disambiguating Illumination and Materials using Unintended Shadows (CVPR 2024 · 6 citations) |
| Remove auxiliary depth loss (pseudo depth from Depth Anything V2) | Pix3D · FS@τ | 35.9 vs 39.6 full | Hurt | Robust 3D Shape Reconstruction in Zero-Shot from a Single Image in the Wild (CVPR 2025 · 2 citations) |
| Add the contrastive encoder and touch shape fusion module | ABC, 5 random grasps · CD | 4.430 -> 3.134 | Helped | Touch2Shape: Touch-Conditioned 3D Diffusion for Shape Exploration and Reconstruction (CVPR 2025 · 2 citations) |
| Use ViT-S instead of the proposed model size | YCB-Video · CD (mm) | 3.56 -> 3.70 | Hurt | RaySt3R: Predicting Novel Depth Maps for Zero-Shot Object Completion (NeurIPS 2025 · 1 citation) |
| add convex/concave flip data augmentation during training | Multistable perception stimuli: four circles, nested rings, star, and snake · Wasserstein distance | 33.79 -> 12.59; 39.56 -> 29.96; 22.04 -> 17.32; 23.02 -> 18.27 | Helped | Multistable Shape from Shading Emerges from Patch Diffusion (NeurIPS 2024 · 1 citation) |
| Add PDFs to the CST base model, with MLP-predicted object and part latent vectors | ShapeNetPart · Chamfer-L1 | 0.239 -> 0.228 | Helped | Learning Partonomic 3D Reconstruction from Image Collections (CVPR 2025 · 0 citations) |
| add gradient-based refinement | ScanNet, all categories · Chamfer distance [m] | 0.14 -> 0.11 | Helped | PyTorchGeoNodes: Enabling Differentiable Shape Programs for 3D Shape Reconstruction (CVPR 2025 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 9 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- ScanNet · 2 of 9 papers
- ShapeNet · 2 of 9 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 38 papers; from OpenAlex and the papers' affiliations).
- Google · 3 papers
- Inria · 2 papers
- Meta · 2 papers
- Georgia Institute of Technology · 2 papers
- University of Illinois Chicago · 2 papers
- Intel · 2 papers
Use it
Download BibTeX of 40 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
- Primitive-based shape abstraction (99 papers)
- Articulated and non-rigid shape learning (75 papers)
- 3D shape generative modeling (74 papers)
- Optimization-based shape matching (59 papers)
Elsewhere in the topic
- Neural implicit surface reconstruction (131 papers)
- Gaussian and implicit head avatars (115 papers)
- Textured mesh generation (111 papers)
- Face reflectance and relightable rendering (85 papers)
- 3D morphable face model regression (80 papers)
- Full-body avatar reconstruction (78 papers)
- Learned mesh and geometric representations (73 papers)
- Clothed human modeling (60 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).