NeurIPS 2024 accepted papers on 3D vision

217 of the 4,560 papers accepted at NeurIPS 2024 (4.8%) are on 3D vision, compared with 3.3% at NeurIPS 2023. They fall into 6 topics.

See NeurIPS 2024 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 · 68 papers

CAT3D: Create Anything in 3D with Multi-View Diffusion Models (39 citations)
Chat-Scene: Bridging 3D Scene and Large Language Models with Object Identifiers (31 citations)
RadarOcc: Robust 3D Occupancy Prediction with 4D Imaging Radar (28 citations)
MAN TruckScenes: A multimodal dataset for autonomous trucking in diverse conditions (18 citations)
Unique3D: High-Quality and Efficient 3D Mesh Generation from a Single Image (18 citations)
DreamScene4D: Dynamic Multi-Object Scene Generation from Monocular Videos (15 citations)
OctreeOcc: Efficient and Multi-Granularity Occupancy Prediction Using Octree Queries (15 citations)
SceneCraft: Layout-Guided 3D Scene Generation (12 citations)
CryoBench: Diverse and challenging datasets for the heterogeneity problem in cryo-EM (11 citations)
Director3D: Real-world Camera Trajectory and 3D Scene Generation from Text (10 citations)
Large Spatial Model: End-to-end Unposed Images to Semantic 3D (8 citations)
SCube: Instant Large-Scale Scene Reconstruction using VoxSplats (8 citations)

and 56 more in the interactive map.

Gaussian splatting · 65 papers

LightGaussian: Unbounded 3D Gaussian Compression with 15x Reduction and 200+ FPS (174 citations)
3D Gaussian Splatting as Markov Chain Monte Carlo (54 citations)
DOGS: Distributed-Oriented Gaussian Splatting for Large-Scale 3D Reconstruction Via Gaussian Consensus (39 citations)
WildGaussians: 3D Gaussian Splatting In the Wild (34 citations)
ContextGS : Compact 3D Gaussian Splatting with Anchor Level Context Model (28 citations)
3DGS-Enhancer: Enhancing Unbounded 3D Gaussian Splatting with View-consistent 2D Diffusion Priors (25 citations)
VCR-GauS: View Consistent Depth-Normal Regularizer for Gaussian Surface Reconstruction (25 citations)
GSDF: 3DGS Meets SDF for Improved Neural Rendering and Reconstruction (24 citations)
L4GM: Large 4D Gaussian Reconstruction Model (23 citations)
Spec-Gaussian: Anisotropic View-Dependent Appearance for 3D Gaussian Splatting (23 citations)
Binocular-Guided 3D Gaussian Splatting with View Consistency for Sparse View Synthesis (21 citations)
HumanSplat: Generalizable Single-Image Human Gaussian Splatting with Structure Priors (19 citations)

and 53 more in the interactive map.

Face and shape · 30 papers

MeshXL: Neural Coordinate Field for Generative 3D Foundation Models (20 citations)
PhyRecon: Physically Plausible Neural Scene Reconstruction (13 citations)
Codec Avatar Studio: Paired Human Captures for Complete, Driveable, and Generalizable Avatars (9 citations)
Expressive Gaussian Human Avatars from Monocular RGB Video (6 citations)
Meta 3D AssetGen: Text-to-Mesh Generation with High-Quality Geometry, Texture, and PBR Materials (5 citations)
DMesh: A Differentiable Mesh Representation (4 citations)
MeshFormer : High-Quality Mesh Generation with 3D-Guided Reconstruction Model (4 citations)
NeuralClothSim: Neural Deformation Fields Meet the Thin Shell Theory (3 citations)
ID-to-3D: Expressive ID-guided 3D Heads via Score Distillation Sampling (2 citations)
Learning 3D Garment Animation from Trajectories of A Piece of Cloth (2 citations)
Neural Experts: Mixture of Experts for Implicit Neural Representations (2 citations)
AlphaTablets: A Generic Plane Representation for 3D Planar Reconstruction from Monocular Videos (1 citation)

and 18 more in the interactive map.

Human pose estimation · 25 papers

Hamba: Single-view 3D Hand Reconstruction with Graph-guided Bi-Scanning Mamba (18 citations)
Neural Localizer Fields for Continuous 3D Human Pose and Shape Estimation (17 citations)
Harmony4D: A Video Dataset for In-The-Wild Close Human Interactions (11 citations)
Point Cloud Matters: Rethinking the Impact of Different Observation Spaces on Robot Learning (10 citations)
A Benchmark Dataset for Event-Guided Human Pose Estimation and Tracking in Extreme Conditions (6 citations)
Posture-Informed Muscular Force Learning for Robust Hand Pressure Estimation (6 citations)
Continuous Heatmap Regression for Pose Estimation via Implicit Neural Representation (5 citations)
emg2pose: A Large and Diverse Benchmark for Surface Electromyographic Hand Pose Estimation (5 citations)
EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity (3 citations)
ManiPose: Manifold-Constrained Multi-Hypothesis 3D Human Pose Estimation (3 citations)
Referring Human Pose and Mask Estimation In the Wild (3 citations)
EfficientCAPER: An End-to-End Framework for Fast and Robust Category-Level Articulated Object Pose Estimation (2 citations)

and 13 more in the interactive map.

Depth and stereo · 19 papers

Depth Anywhere: Enhancing 360 Monocular Depth Estimation via Perspective Distillation and Unlabeled Data Augmentation (12 citations)
Virtual Scanning: Unsupervised Non-line-of-sight Imaging from Irregularly Undersampled Transients (10 citations)
DCDepth: Progressive Monocular Depth Estimation in Discrete Cosine Domain (8 citations)
SfPUEL: Shape from Polarization under Unknown Environment Light (8 citations)
A Simple yet Universal Framework for Depth Completion (3 citations)
Metric from Human: Zero-shot Monocular Metric Depth Estimation via Test-time Adaptation (3 citations)
Zero-Shot Event-Intensity Asymmetric Stereo via Visual Prompting from Image Domain (3 citations)
RobIR: Robust Inverse Rendering for High-Illumination Scenes (2 citations)
RSA: Resolving Scale Ambiguities in Monocular Depth Estimators through Language Descriptions (2 citations)
SS3DM: Benchmarking Street-View Surface Reconstruction with a Synthetic 3D Mesh Dataset (2 citations)
$SE(3)$ Equivariant Ray Embeddings for Implicit Multi-View Depth Estimation (1 citation)
LoD-Loc: Aerial Visual Localization using LoD 3D Map with Neural Wireframe Alignment (1 citation)

and 7 more in the interactive map.