ICCV 2021 accepted papers on 3D vision
291 of the 1,612 papers accepted at ICCV 2021 (18.1%) are on 3D vision, compared with 19.4% at ICCV 2019. They fall into 6 topics.
See ICCV 2021 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
Human pose estimation · 90 papers
- 3D Human Pose Estimation With Spatial and Temporal Transformers (778 citations)
- TransPose: Keypoint Localization via Transformer (493 citations)
- TokenPose: Learning Keypoint Tokens for Human Pose Estimation (472 citations)
- PARE: Part Attention Regressor for 3D Human Body Estimation (458 citations)
- Mesh Graphormer (367 citations)
- PyMAF: 3D Human Pose and Shape Regression With Pyramidal Mesh Alignment Feedback Loop (359 citations)
- HuMoR: 3D Human Motion Model for Robust Pose Estimation (346 citations)
- MSR-GCN: Multi-Scale Residual Graph Convolution Networks for Human Motion Prediction (286 citations)
- Human Pose Regression With Residual Log-Likelihood Estimation (282 citations)
- Monocular, One-Stage, Regression of Multiple 3D People (277 citations)
- H2O: Two Hands Manipulating Objects for First Person Interaction Recognition (214 citations)
- Modulated Graph Convolutional Network for 3D Human Pose Estimation (204 citations)
and 78 more in the interactive map.
Depth and stereo · 88 papers
- iMAP: Implicit Mapping and Positioning in Real-Time (761 citations)
- Revisiting Stereo Depth Estimation From a Sequence-to-Sequence Perspective With Transformers (354 citations)
- COTR: Correspondence Transformer for Matching Across Images (314 citations)
- Omnidata: A Scalable Pipeline for Making Multi-Task Mid-Level Vision Datasets From 3D Scans (291 citations)
- AA-RMVSNet: Adaptive Aggregation Recurrent Multi-View Stereo Network (216 citations)
- Pixel-Perfect Structure-From-Motion With Featuremetric Refinement (205 citations)
- R-MSFM: Recurrent Multi-Scale Feature Modulation for Monocular Depth Estimating (189 citations)
- Learning Multi-Scene Absolute Pose Regression With Transformers (160 citations)
- EPP-MVSNet: Epipolar-Assembling Based Depth Prediction for Multi-View Stereo (157 citations)
- Single-Shot Hyperspectral-Depth Imaging With Learned Diffractive Optics (142 citations)
- Fine-Grained Semantics-Aware Representation Enhancement for Self-Supervised Monocular Depth Estimation (140 citations)
- Multi-View 3D Reconstruction With Transformers (128 citations)
and 76 more in the interactive map.
Face and shape · 58 papers
- UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction (682 citations)
- Common Objects in 3D: Large-Scale Learning and Evaluation of Real-Life 3D Category Reconstruction (411 citations)
- Fake It Till You Make It: Face Analysis in the Wild Using Synthetic Data Alone (286 citations)
- MVTN: Multi-View Transformation Network for 3D Shape Recognition (265 citations)
- MeshTalk: 3D Face Animation From Speech Using Cross-Modality Disentanglement (259 citations)
- SNARF: Differentiable Forward Skinning for Animating Non-Rigid Neural Implicit Shapes (224 citations)
- ARCH++: Animation-Ready Clothed Human Reconstruction Revisited (213 citations)
- Neural Articulated Radiance Field (183 citations)
- Modulated Periodic Activations for Generalizable Local Functional Representations (134 citations)
- HeadGAN: One-Shot Neural Head Synthesis and Editing (130 citations)
- DeepMultiCap: Performance Capture of Multiple Characters Using Sparse Multiview Cameras (124 citations)
- Learning Signed Distance Field for Multi-View Surface Reconstruction (117 citations)
and 46 more in the interactive map.
NeRF and radiance fields · 29 papers
- Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields (2,200 citations)
- Nerfies: Deformable Neural Radiance Fields (1,284 citations)
- PlenOctrees for Real-Time Rendering of Neural Radiance Fields (928 citations)
- MVSNeRF: Fast Generalizable Radiance Field Reconstruction From Multi-View Stereo (902 citations)
- KiloNeRF: Speeding Up Neural Radiance Fields With Thousands of Tiny MLPs (768 citations)
- BARF: Bundle-Adjusting Neural Radiance Fields (689 citations)
- FastNeRF: High-Fidelity Neural Rendering at 200FPS (631 citations)
- Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis (511 citations)
- Baking Neural Radiance Fields for Real-Time View Synthesis (500 citations)
- NeRD: Neural Reflectance Decomposition From Image Collections (491 citations)
- Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synthesis of a Dynamic Scene From Monocular Video (477 citations)
- AD-NeRF: Audio Driven Neural Radiance Fields for Talking Head Synthesis (438 citations)
and 17 more in the interactive map.
3D generation · 24 papers
- AI Choreographer: Music Conditioned 3D Dance Generation With AIST++ (706 citations)
- Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understanding (446 citations)
- 3D-FRONT: 3D Furnished Rooms With layOuts and semaNTics (336 citations)
- FIERY: Future Instance Prediction in Bird's-Eye View From Surround Monocular Cameras (297 citations)
- Hand-Object Contact Consistency Reasoning for Human Grasps Generation (190 citations)
- Synthesis of Compositional Animations From Textual Descriptions (186 citations)
- Infinite Nature: Perpetual View Generation of Natural Scenes From a Single Image (176 citations)
- Stochastic Scene-Aware Motion Prediction (175 citations)
- Reconstructing Hand-Object Interactions in the Wild (163 citations)
- Structured Bird's-Eye-View Traffic Scene Understanding From Onboard Images (122 citations)
- Worldsheet: Wrapping the World in a 3D Sheet for View Synthesis From a Single Image (79 citations)
- VMNet: Voxel-Mesh Network for Geodesic-Aware 3D Semantic Segmentation (76 citations)
and 12 more in the interactive map.
