NeurIPS 2020 accepted papers on 3D vision
39 of the 1,901 papers accepted at NeurIPS 2020 (2.1%) are on 3D vision, compared with 1.7% at NeurIPS 2019. They fall into 5 topics.
See NeurIPS 2020 in the interactive map
Submitting to NeurIPS? 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 · 20 papers
- Multiview Neural Surface Reconstruction by Disentangling Geometry and Appearance (145 citations)
- Geo-PIFu: Geometry and Pixel Aligned Implicit Functions for Single-view Human Reconstruction (116 citations)
- Generative 3D Part Assembly via Dynamic Graph Learning (49 citations)
- UCSG-NET- Unsupervised Discovering of Constructive Solid Geometry Tree (40 citations)
- 3D Shape Reconstruction from Vision and Touch (29 citations)
- Convolutional Generation of Textured 3D Meshes (29 citations)
- ShapeFlow: Learnable Deformation Flows Among 3D Shapes (28 citations)
- MeshSDF: Differentiable Iso-Surface Extraction (22 citations)
- Neural Non-Rigid Tracking (22 citations)
- Online Adaptation for Consistent Mesh Reconstruction in the Wild (18 citations)
- LoopReg: Self-supervised Learning of Implicit Surface Correspondences, Pose and Shape for 3D Human Mesh Registration (17 citations)
- SDF-SRN: Learning Signed Distance 3D Object Reconstruction from Static Images (14 citations)
and 8 more in the interactive map.
3D generation · 9 papers
- BlockGAN: Learning 3D Object-aware Scene Representations from Unlabelled Images (108 citations)
- Partially View-aligned Clustering (46 citations)
- RELATE: Physically Plausible Multi-Object Scene Synthesis Using Structured Latent Spaces (35 citations)
- Learning Object-Centric Representations of Multi-Object Scenes from Multiple Views (18 citations)
- Continuous Object Representation Networks: Novel View Synthesis without Target View Supervision (11 citations)
- Multi-Plane Program Induction with 3D Box Priors (5 citations)
- Generative View Synthesis: From Single-view Semantics to Novel-view Images (4 citations)
- Rel3D: A Minimally Contrastive Benchmark for Grounding Spatial Relations in 3D (4 citations)
- Unsupervised object-centric video generation and decomposition in 3D (4 citations)
Depth and stereo · 5 papers
- An Analysis of SVD for Deep Rotation Estimation (41 citations)
- HyNet: Learning Local Descriptor with Hybrid Similarity Measure and Triplet Loss (12 citations)
- Displacement-Invariant Matching Cost Learning for Accurate Optical Flow Estimation (9 citations)
- GOCor: Bringing Globally Optimized Correspondence Volumes into Your Neural Network (7 citations)
- Forget About the LiDAR: Self-Supervised Depth Estimators with MED Probability Volumes
Human pose estimation · 3 papers
NeRF and radiance fields · 2 papers
- GRAF: Generative Radiance Fields for 3D-Aware Image Synthesis (319 citations)
- Neural Sparse Voxel Fields (11 citations)
