NeurIPS 2026 accepted papers on 3D vision
246 of the 6,080 papers accepted at NeurIPS 2026 (4.0%) are on 3D vision, compared with 4.6% at NeurIPS 2025. They fall into 6 topics.
See NeurIPS 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
3D generation · 85 papers
- 3DCodeBench: Benchmarking Agentic Procedural 3D Modeling Via Code
- AdaOcc: Adaptive 3D Occupancy Prediction for Embodied Tasks
- At FullTilt: Real-Time Open-Set 3D Macromolecule Detection Directly from Tilted 2D Projections
- CAIRN: Cross-Room 3D Scene Understanding with Topology-Aware Large Multimodal Models
- Controllable Dynamic 3D Shape Generation via 3D Trajectories and Text
- DecompDreamer: A Composition-Aware Curriculum for Structured 3D Asset Generation
- DrawingsDreamer: A Unified Multi-View Engineering Drawings Generation Model
- Effective Multi-sensor Conditioning for Street-view Novel-view Synthesis
- Enhancing Novel View Synthesis via Geometry Grounded Set Diffusion
- FlowHOI: Flow-based Semantics-Grounded Generation of Hand-Object Interactions for Dexterous Robot Manipulation
- GenRec: Knowing Where to Reconstruct and Where to Generate
- GeoWind2Plan: Mission-Time 3D Urban Wind Prediction for Energy-Efficient UAV Planning
and 73 more in the interactive map.
Gaussian splatting · 54 papers
- 3D Point Splatting for mmWave Radar Novel View Synthesis
- 3D Skew Normal Splatting
- AdpSplit: Error-Driven Adaptive Splitting for Faster Geometry Discovery in 3D Gaussian Splatting
- BEAST3D: animal behavioral analysis and neural encoding from multi-view video via Gaussian splatting
- Coarse-to-Real: Generative Rendering for Populated Dynamic Scenes
- CodecSplat: Ultra-Compact Latent Coding for Feed-Forward 3D Gaussian Splatting
- DySurface: Consistent 4D Surface Reconstruction via Bridging Explicit Gaussians and Implicit Functions
- FRUC: Feedforward Dynamic Scene Reconstruction from Uncalibrated Collaborative Driving Views
- Gaussian Splatting-based Volumetric Video Compression with Sparse 4D Anchors
- GLUT: 3D Gaussian Lookup Table for Continuous Color Transformation
- MACRO: Training-free Multi-plane Attention for Closeup Render Optimization
- One World, Dual Timeline: Decoupled Spatio-Temporal Gaussian Scene Graph for 4D Cooperative Driving Reconstruction
and 42 more in the interactive map.
Depth and stereo · 42 papers
- BA-T: An Iterative Transformer for Two-View Bundle Adjustment
- Beyond Row Alignment: Virtual-Camera-Aware Online Stereo Rectification
- Change-Robust Online Topological Memory for Long-Term Relocalization and Semantic Navigation
- CLoSeR: Closing the Loop for Long-Context Streaming Reconstruction
- CM-EVS: Sparse Panoramic RGB-D-Pose Data for Complete Scene Coverage
- DarkVGGT: Seeing Through Darkness Using Thermal Geometry without Daylight Tax
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth
- DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images
- DPPE: Rethinking Camera-Based Positional Encoding for Scaling Multi-View Transformers
- Déjà View: Looping Transformers for Multi-View 3D Reconstruction
- H-Flow: Self-supervised Human Scene Flow via Physics-inspired Joint Multi-modal Learning
- Less Decoder is More Encoder: Geometric Representation Learning from Novel View Synthesis
and 30 more in the interactive map.
Face and shape · 31 papers
- BrickAnything: Geometry-Conditioned Buildable Brick Generation with Structure-Aware Tokenization
- DiffeoMorph: Learning to Morph 3D Shapes Using Differentiable Agent-Based Simulations
- DinoComplete: 3D Shape Completion with Distilled Semantic Priors and State Space Models
- DressWild: Feed-Forward Pose-Agnostic Garment Sewing Pattern Generation from In-the-Wild Images
- FA-LAM: Focus-Aware Large Avatar Model for One-Shot 4D Animatable Gaussian Head
- Flow-Corrected Shape Optimization: Taming Manifold Drift in High-Dimensional 3D Models
- M-plicits: Neural Implicit Surfaces via Nested Multiscale Residuals
- Mesh BDF: Barycentric Dominance Field for 3D Native Mesh Generation
- Pygmalion Effect in Vision: Image-to-Clay Translation for Reflective Geometry Reconstruction
- Rigel3D: Rig-aware Latents for Animation-Ready 3D Asset Generation
- S2MDF: A Plug-And-Play Layer for Intersection-Free Multi-Object Signed Distance Fields
- SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild
and 19 more in the interactive map.
Human pose estimation · 27 papers
- Articulation in Prime: Primitive-Based Articulated Object Understanding from a Single Casual Video
- COMPOSE: Hypergraph Cover Optimization for Multi-view 3D Human Pose Estimation
- ComPose: When to Trust Hands for Object Pose Tracking
- Emergence of a Shared Canonical Object Frame from In-the-Wild Videos
- FactorizedHMR: A Hybrid Framework for Video Human Mesh Recovery
- GenCOPE: Syn2Real Generalized Category-Level Object Pose Estimation for Robotic Picking
- HOPE: Hand-Object Pressure Estimation from Monocular Videos
- Neural-Behavioral Representation of Natural Whole-body Movement in Monkeys
- PDF-HR: Pose Distance Fields for Humanoid Robots
- ProxyPose: 6-DoF Pose Tracking via Video-to-Video Translation
- SimpliHuMoN: Simplifying Human Motion Prediction
- UniCon3R: Unified Contact-aware 4D Human-Scene Reconstruction from Monocular Video
and 15 more in the interactive map.
NeRF and radiance fields · 7 papers
- CD-RCM: Generalizable Continuous-Depth Novel View Synthesis for Reflectance Confocal Microscopy
- Resolving Representation Ambiguity in Feedforward Novel View Synthesis Transformer via Semantic-Spatial Decoupling
- NeRFix: fixing subtle mistakes in the quadrature of the NeRF volumetric integral
- RASS: Risk-Audited Budget Selection for Compact NeRF Benchmark Subsets
- SCION: Scene Composition via Instanced Neural Primitives
- Spik-NeRF v2: Pushing the Limit of Spiking Neural Radiance Fields with $ \pm $I-LIF
- TTB: Test-time MLP Baking for Efficient Rendering of Decoder-only View Synthesis Models