Research map / 3D vision / 3D generation
Feed-forward 3D asset generation and editing
Generates, textures and edits 3D content using reconstruction models and decomposed outputs.
116 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2022–2026) · an approach in 3D generation, keyword cluster “3d generation · text · consistent”
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.
31 papers in 2023–24 and 81 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 6.8% to 9.2%.
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
- 2022 · 3D Photo Stylization: Learning To Generate Stylized Novel Views From a Single Image (CVPR 2022 · 46 citations)
- 2023 · NeuralLift-360: Lifting an In-the-Wild 2D Photo to a 3D Object With 360deg Views (CVPR 2023 · 85 citations)
- 2024 · AvatarVerse: High-Quality & Stable 3D Avatar Creation from Text and Pose (AAAI 2024 · 45 citations)
- 2025 · Enhanced Contrastive Learning with Multi-view Longitudinal Data for Chest X-ray Report Generation (CVPR 2025 · 25 citations)
- 2026 · LATTICE: Democratize High-Fidelity 3D Generation at Scale (CVPR 2026 · 2 citations)
Earliest papers here
- 3D Photo Stylization: Learning To Generate Stylized Novel Views From a Single Image (CVPR 2022 · 46 citations)
- ME-GAN: Learning Panoptic Electrocardio Representations for Multi-view ECG Synthesis Conditioned on Heart Diseases (ICML 2022 · 10 citations)
- Pose Guided Image Generation from Misaligned Sources via Residual Flow Based Correction (AAAI 2022 · 5 citations)
What has been tried
68 papers here report 750 measured changes (a component added, removed or swapped, compared with a baseline): helped 77%, no clear effect 3%, hurt 18%, 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 progressive grid | Progressive-strategy ablation · visual detail | Helped | AvatarVerse: High-Quality & Stable 3D Avatar Creation from Text and Pose (AAAI 2024 · 45 citations) | |
| Add relative depth ranking supervision | 10 scene completion scenes · Corrs | 682 -> 696 | Helped | NeRFiller: Completing Scenes via Generative 3D Inpainting (CVPR 2024 · 30 citations) |
| Add multi-view longitudinal contrastive learning | MIMIC-CXR · BLEU-4 | 0.136 -> 0.158 | Helped | Enhanced Contrastive Learning with Multi-view Longitudinal Data for Chest X-ray Report Generation (CVPR 2025 · 25 citations) |
| Add cosine noise scheduling | Unconditional hairstyle generation · FID | 62.1 -> 36.9 | Helped | Text-Conditioned Generative Model of 3D Strand-based Human Hairstyles (CVPR 2024 · 21 citations) |
| Use “realistic Sc texture in top view” instead of generic generation prompts | ACG material generation · FID | 1.89 -> 1.55 (albedo); 1.74 -> 1.27 (normal); 1.83 -> 1.59 (roughness) | Helped | Material Palette: Extraction of Materials from a Single Image (CVPR 2024 · 19 citations) |
| Add MF-Attention | Ablation study · PFC | 0.9600 -> 0.9290 | Helped | POPDG: Popular 3D Dance Generation with PopDanceSet (CVPR 2024 · 19 citations) |
| add expression deformation maps | Nersemble self-reenactment, 10 reference images · PSNR | 21.64 -> 22.54 | Helped | CAP4D: Creating Animatable 4D Portrait Avatars with Morphable Multi-View Diffusion Models (CVPR 2025 · 17 citations) |
| Add global depth regularization | Global editing · Structure Dist. | 0.138 -> 0.008 | Helped | SHAP-EDITOR: Instruction-Guided Latent 3D Editing in Seconds (CVPR 2024 · 15 citations) |
| Add Stage 1 generative pre-training | Chair [39] · FID | 25.0112 -> 8.7701 | Helped | Rapid 3D Model Generation with Intuitive 3D Input (CVPR 2024 · 13 citations) |
| Replace RGB-D VAE with RGB VAE | Video generation model · FID | 22.5526 -> 26.1607 | Hurt | GenFusion: Closing the Loop between Reconstruction and Generation via Videos (CVPR 2025 · 10 citations) |
| Use separate 3D bounding-box encoding rather than encoding road and object semantics together in the BEV map | nuScenes val, BEV segmentation · vehicle mIoU | Helped | MagicDrive: Street View Generation with Diverse 3D Geometry Control (ICLR 2024 · 10 citations) | |
| add 3D Gaussian geometric conditioning | AnimeRig test set · IoU | 56.56% -> 70.48% | Helped | DRiVE: Diffusion-based Rigging Empowers Generation of Versatile and Expressive Characters (CVPR 2025 · 9 citations) |
What reviewers will expect
Datasets and baselines named in the 42 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- Objaverse · 5 of 42 papers
- FFHQ · 3 of 42 papers
- ShapeNet · 2 of 42 papers
- MEAD · 2 of 42 papers
- Tanks and Temples · 2 of 42 papers
- GSO · 2 of 42 papers
- Toys4K · 2 of 42 papers
- Poly Haven HDR environment maps · 2 of 42 papers
Baselines
- LGM · 4 of 42 papers
- InstantMesh · 3 of 42 papers
- OpenLRM · 3 of 42 papers
- Magic123 · 2 of 42 papers
- Wonder3D · 2 of 42 papers
- RigNet · 2 of 42 papers
- CharacterGen · 2 of 42 papers
- CLIP-Score · 2 of 42 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 112 papers; from OpenAlex and the papers' affiliations).
- Shanghai Jiao Tong University · 5 papers
- The Chinese University of Hong Kong · 5 papers
- Tsinghua University · 4 papers
- Hong Kong University of Science and Technology · 4 papers
- Shanghai AI Laboratory · 4 papers
- Zhejiang University · 4 papers
- NVIDIA · 4 papers
- National University of Singapore · 4 papers
- University of Science and Technology of China · 4 papers
- Nanyang Technological University · 3 papers
Use it
Download BibTeX of 28 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
- Diffusion-based novel view synthesis (156 papers)
- Score distillation for text-to-3D (106 papers)
- Text-to-3D scene generation (58 papers)
Elsewhere in the topic
- Learning 3D reconstruction from images (196 papers)
- Semantic scene completion (153 papers)
- Generative indoor scene synthesis (131 papers)
- Hand-object interaction generation (116 papers)
- Open-vocabulary 3D scene understanding (95 papers)
- 2D-to-3D pretrained representations (94 papers)
- Monocular 3D object reconstruction (94 papers)
- Scene-aware human motion synthesis (82 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).