Gaussian splatting: research map
1,340 papers on Gaussian splatting in 3D vision, from ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR (2016–2026), grouped into 2 clusters and 7 approaches. The busiest year so far is 2026.
What it is
Gaussian splatting represents a 3D scene as millions of small, coloured, semi-transparent blobs (3D Gaussians) fitted to a set of photos. Once fitted, the scene can be shown from any new viewpoint in real time, much faster than NeRF. Current work makes these scenes smaller and sharper, builds them from only a few photos, and extends them to moving scenes and animatable people.
Terms you will see on this page:
- 3DGS
- 3D Gaussian splatting, the method behind this topic.
- Novel view synthesis
- rendering a scene from a camera position that was not photographed.
- Primitive
- one Gaussian blob; a scene is made of very many of them.
- 4D / dynamic scene
- a scene that changes over time, modelled by moving or deforming the Gaussians.
Start reading with the most cited papers here:
- 4D Gaussian Splatting for Real-Time Dynamic Scene Rendering (CVPR 2024 · 737 citations)
- Mip-Splatting: Alias-free 3D Gaussian Splatting (CVPR 2024 · 524 citations)
Within 3D vision, its share grew from 6.9% in 2023–24 to 26.0% in 2025–26 (122 → 719 papers at ICML, NeurIPS, ICLR, CVPR and AAAI, the venues with data for all four years).
Explore Gaussian splatting in the interactive map
Working on something in this topic? Describe your idea in scime atlas to see which approach it falls under, the closest papers by meaning and how crowded the spot has become.
What changed recently
Approaches whose share of the topic moved most between 2023–24 and 2025–26 (ICML, NeurIPS, ICLR, CVPR and AAAI):
- Compact and efficient Gaussian splatting rising: 24% → 33% of the topic (29 → 240 papers)
- Sparse-view Gaussian splatting for novel views rising: 11% → 14% of the topic (13 → 98 papers)
- Animatable Gaussian human avatars falling: 9% → 4% of the topic (11 → 31 papers)
- Deformable 4D Gaussian splatting falling: 16% → 13% of the topic (19 → 93 papers)
Papers from 2025–2026 that are already cited most:
- DIFIX3D+: Improving 3D Reconstructions with Single-Step Diffusion Models (CVPR 2025 · 48 citations)
- NexusGS: Sparse View Synthesis with Epipolar Depth Priors in 3D Gaussian Splatting (CVPR 2025 · 24 citations)
Approaches and key papers
Static scenes: quality, size and sparse views · 1,043 papers
Keywords: 3d gaussian, gaussian splatting, splatting 3dgs
Approaches in this cluster:
- Compact and efficient Gaussian splatting (400 papers)
Prunes, sparsifies and optimizes Gaussian primitives for faster, smaller 3DGS rendering. - Regularization and primitive design for 3DGS (324 papers)
Improves splatting quality through regularization, new kernels and rendering optimizations. - Sparse-view Gaussian splatting for novel views (257 papers)
Uses depth, epipolar and view priors for generalizable novel view synthesis from few images. - Animatable Gaussian human avatars (62 papers)
Builds deformable Gaussian representations of humans and hands for real-time avatar rendering.
Most cited and most cited since 2024:
- Mip-Splatting: Alias-free 3D Gaussian Splatting (CVPR 2024 · 524 citations)
- SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering (CVPR 2024 · 478 citations)
- Scaffold-GS: Structured 3D Gaussians for View-Adaptive Rendering (CVPR 2024 · 463 citations)
- Gaussian Splatting SLAM (CVPR 2024 · 461 citations)
Dynamic and 4D scenes · 297 papers
Keywords: dynamic, 4d, motion
Approaches in this cluster:
- Deformable 4D Gaussian splatting (138 papers)
Models dynamic scenes with deformable or explicit motion Gaussians for real-time rendering. - Monocular dynamic scene reconstruction (142 papers)
Reconstructs and generates 4D scenes from monocular video using free-moving Gaussian primitives. - Animatable Gaussian human avatars (17 papers)
Represents animatable humans and heads with pose-dependent or mesh-embedded Gaussians.
Most cited and most cited since 2024:
- 4D Gaussian Splatting for Real-Time Dynamic Scene Rendering (CVPR 2024 · 737 citations)
- Deformable 3D Gaussians for High-Fidelity Monocular Dynamic Scene Reconstruction (CVPR 2024 · 420 citations)
- DrivingGaussian: Composite Gaussian Splatting for Surrounding Dynamic Autonomous Driving Scenes (CVPR 2024 · 223 citations)
- Spacetime Gaussian Feature Splatting for Real-Time Dynamic View Synthesis (CVPR 2024 · 186 citations)
Related topics in 3D vision
- Human pose estimation (2,063)
- NeRF and radiance fields (733)
- 3D generation (2,448)
- Face and shape (2,246)
- Depth and stereo (2,574)
