ICCV 2023 accepted papers on 3D vision
450 of the 2,156 papers accepted at ICCV 2023 (20.9%) are on 3D vision, compared with 18.1% at ICCV 2021. They fall into 6 topics.
See ICCV 2023 in the interactive map
Submitting to ICCV? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Face and shape · 95 papers
- NeuS2: Fast Learning of Neural Implicit Surfaces for Multi-view Reconstruction (333 citations)
- EmoTalk: Speech-Driven Emotional Disentanglement for 3D Face Animation (192 citations)
- Delicate Textured Mesh Recovery from NeRF via Adaptive Surface Refinement (168 citations)
- SHERF: Generalizable Human NeRF from a Single Image (108 citations)
- NeRF-LOAM: Neural Implicit Representation for Large-Scale Incremental LiDAR Odometry and Mapping (103 citations)
- AvatarCraft: Transforming Text into Neural Human Avatars with Parameterized Shape and Pose Control (92 citations)
- DNA-Rendering: A Diverse Neural Actor Repository for High-Fidelity Human-Centric Rendering (91 citations)
- NeuRBF: A Neural Fields Representation with Adaptive Radial Basis Functions (86 citations)
- AG3D: Learning to Generate 3D Avatars from 2D Image Collections (69 citations)
- Ref-NeuS: Ambiguity-Reduced Neural Implicit Surface Learning for Multi-View Reconstruction with Reflection (69 citations)
- DiFaReli: Diffusion Face Relighting (56 citations)
- One-shot Implicit Animatable Avatars with Model-based Priors (54 citations)
and 83 more in the interactive map.
Human pose estimation · 92 papers
- MotionBERT: A Unified Perspective on Learning Human Motion Representations (439 citations)
- Humans in 4D: Reconstructing and Tracking Humans with Transformers (344 citations)
- Diffusion-Based 3D Human Pose Estimation with Multi-Hypothesis Aggregation (178 citations)
- GLA-GCN: Global-local Adaptive Graph Convolutional Network for 3D Human Pose Estimation from Monocular Video (167 citations)
- PoseDiffusion: Solving Pose Estimation via Diffusion-aided Bundle Adjustment (115 citations)
- SiLK: Simple Learned Keypoints (112 citations)
- GPGait: Generalized Pose-based Gait Recognition (111 citations)
- UniDexGrasp++: Improving Dexterous Grasping Policy Learning via Geometry-Aware Curriculum and Iterative Generalist-Specialist Learning (109 citations)
- DiffPose: Multi-hypothesis Human Pose Estimation using Diffusion Models (99 citations)
- ActFormer: A GAN-based Transformer towards General Action-Conditioned 3D Human Motion Generation (92 citations)
- Full-Body Articulated Human-Object Interaction (75 citations)
- EMDB: The Electromagnetic Database of Global 3D Human Pose and Shape in the Wild (72 citations)
and 80 more in the interactive map.
3D generation · 91 papers
- Zero-1-to-3: Zero-shot One Image to 3D Object (1,133 citations)
- Fantasia3D: Disentangling Geometry and Appearance for High-quality Text-to-3D Content Creation (577 citations)
- ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenes (434 citations)
- SurroundOcc: Multi-camera 3D Occupancy Prediction for Autonomous Driving (332 citations)
- Make-It-3D: High-fidelity 3D Creation from A Single Image with Diffusion Prior (315 citations)
- OccFormer: Dual-path Transformer for Vision-based 3D Semantic Occupancy Prediction (288 citations)
- OpenOccupancy: A Large Scale Benchmark for Surrounding Semantic Occupancy Perception (250 citations)
- Text2Tex: Text-driven Texture Synthesis via Diffusion Models (248 citations)
- Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models (246 citations)
- Generative Novel View Synthesis with 3D-Aware Diffusion Models (211 citations)
- Single-Stage Diffusion NeRF: A Unified Approach to 3D Generation and Reconstruction (160 citations)
- Scene as Occupancy (153 citations)
and 79 more in the interactive map.
Depth and stereo · 87 papers
- LightGlue: Local Feature Matching at Light Speed (1,188 citations)
- Sample4Geo: Hard Negative Sampling For Cross-View Geo-Localisation (240 citations)
- GO-SLAM: Global Optimization for Consistent 3D Instant Reconstruction (184 citations)
- PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point Tracking (146 citations)
- CroCo v2: Improved Cross-view Completion Pre-training for Stereo Matching and Optical Flow (137 citations)
- Towards Zero-Shot Scale-Aware Monocular Depth Estimation (122 citations)
- NDDepth: Normal-Distance Assisted Monocular Depth Estimation (96 citations)
- Robust Monocular Depth Estimation under Challenging Conditions (87 citations)
- When Epipolar Constraint Meets Non-Local Operators in Multi-View Stereo (62 citations)
- Self-supervised Monocular Depth Estimation: Let's Talk About The Weather (59 citations)
- Boosting 3-DoF Ground-to-Satellite Camera Localization Accuracy via Geometry-Guided Cross-View Transformer (51 citations)
- ELFNet: Evidential Local-global Fusion for Stereo Matching (47 citations)
and 75 more in the interactive map.
NeRF and radiance fields · 79 papers
- Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields (689 citations)
- Instruct-NeRF2NeRF: Editing 3D Scenes with Instructions (446 citations)
- SparseNeRF: Distilling Depth Ranking for Few-shot Novel View Synthesis (324 citations)
- Tri-MipRF: Tri-Mip Representation for Efficient Anti-Aliasing Neural Radiance Fields (145 citations)
- Efficient Region-Aware Neural Radiance Fields for High-Fidelity Talking Portrait Synthesis (132 citations)
- NerfAcc: Efficient Sampling Accelerates NeRFs (130 citations)
- MatrixCity: A Large-scale City Dataset for City-scale Neural Rendering and Beyond (128 citations)
- NeILF++: Inter-Reflectable Light Fields for Geometry and Material Estimation (88 citations)
- CLIPascene: Scene Sketching with Different Types and Levels of Abstraction (87 citations)
- GRAM-HD: 3D-Consistent Image Generation at High Resolution with Generative Radiance Manifolds (80 citations)
- Nerfbusters: Removing Ghostly Artifacts from Casually Captured NeRFs (72 citations)
- Neural LiDAR Fields for Novel View Synthesis (70 citations)
and 67 more in the interactive map.
Gaussian splatting · 6 papers
- LERF: Language Embedded Radiance Fields (488 citations)
- Mixed Neural Voxels for Fast Multi-view Video Synthesis (120 citations)
- Total-Recon: Deformable Scene Reconstruction for Embodied View Synthesis (16 citations)
- Clutter Detection and Removal in 3D Scenes with View-Consistent Inpainting (11 citations)
- Spacetime Surface Regularization for Neural Dynamic Scene Reconstruction (10 citations)
- 3D Motion Magnification: Visualizing Subtle Motions from Time-Varying Radiance Fields
