TPAMI 2025 papers on 3D vision
144 of the 932 papers published in TPAMI 2025 (15.5%) are on 3D vision, compared with 9.5% at TPAMI 2024. They fall into 6 topics.
See TPAMI 2025 in the interactive map
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
3D generation · 31 papers
- ViewCrafter: Taming Video Diffusion Models for High-fidelity Novel View Synthesis (448 citations)
- V3D: Video Diffusion Models are Effective 3D Generators (111 citations)
- PonderV2: Improved 3D Representation With a Universal Pre-Training Paradigm (62 citations)
- Benchmarking and Improving Bird’s Eye View Perception Robustness in Autonomous Driving (45 citations)
- S-NeRF++: Autonomous Driving Simulation via Neural Reconstruction and Generation (37 citations)
- Compositional Generative Model of Unbounded 4D Cities (22 citations)
- ManiDext: Hand-Object Manipulation Synthesis via Continuous Correspondence Embeddings and Residual-Guided Diffusion (22 citations)
- Gen-3Diffusion: Realistic Image-to-3D Generation Via 2D & 3D Diffusion Synergy (20 citations)
- Planner3D: LLM-enhanced Graph Prior Meets 3D Indoor Scene Explicit Regularization (16 citations)
- PanopticNeRF-360: Panoramic 3D-to-2D Label Transfer in Urban Scenes (11 citations)
- DreamCraft3D++: Efficient Hierarchical 3D Generation with Multi-Plane Reconstruction Model (10 citations)
- DiffTF++: 3D-Aware Diffusion Transformer for Large-Vocabulary 3D Generation (9 citations)
and 19 more in the interactive map.
Depth and stereo · 31 papers
- UniDepthV2: Universal Monocular Metric Depth Estimation Made Simpler (195 citations)
- IGEV++: Iterative Multi-Range Geometry Encoding Volumes for Stereo Matching (112 citations)
- Event-Based Stereo Depth Estimation: A Survey (36 citations)
- Revisiting One-Stage Deep Uncalibrated Photometric Stereo via Fourier Embedding (25 citations)
- SceneTracker: Long-Term Scene Flow Estimation Network (23 citations)
- Flow-Anything: Learning Real-World Optical Flow Estimation From Large-Scale Single-View Images (18 citations)
- MOWA: Multiple-in-One Image Warping Model (15 citations)
- Event Cameras Meet SPADs for High-Speed, Low-Bandwidth Imaging (13 citations)
- Learning Lens Blur Fields (13 citations)
- Investigating Synthetic-to-Real Transfer Robustness for Stereo Matching and Optical Flow Estimation (12 citations)
- Revisiting Gradient-Based Uncertainty for Monocular Depth Estimation (12 citations)
- SAS: A General Framework Induced by Sequence Association for Shape From Focus (10 citations)
and 19 more in the interactive map.
Human pose estimation · 26 papers
- HybrIK-X: Hybrid Analytical-Neural Inverse Kinematics for Whole-Body Mesh Recovery (69 citations)
- GDRNPP: A Geometry-Guided and Fully Learning-Based Object Pose Estimator (56 citations)
- SMPLest-X: Ultimate Scaling for Expressive Human Pose and Shape Estimation (45 citations)
- Affine Correspondences Between Multi-Camera Systems for Relative Pose Estimation (29 citations)
- Diff9D: Diffusion-Based Domain-Generalized Category-Level 9-DoF Object Pose Estimation (29 citations)
- Another Vertical View: A Hierarchical Network for Heterogeneous Trajectory Prediction via Spectrums (16 citations)
- Reconstructing Three-Dimensional Models of Interacting Humans (16 citations)
- High-Resolution Open-Vocabulary Object 6D Pose Estimation (12 citations)
- From FastPoseGait to GPGait++: Bridging the Past and Future for Pose-Based Gait Recognition (11 citations)
- Semi-Supervised Unconstrained Head Pose Estimation in the Wild (9 citations)
- AniMer+: Unified Pose and Shape Estimation Across Mammalia and Aves via Family-Aware Transformer (8 citations)
- Condition Numbers in Multiview Geometry, Instability in Relative Pose Estimation, and RANSAC (6 citations)
and 14 more in the interactive map.
Gaussian splatting · 22 papers
- Octree-GS: Towards Consistent Real-time Rendering with LOD-Structured 3D Gaussians (324 citations)
- HAC++: Towards 100X Compression of 3D Gaussian Splatting (84 citations)
- GIR: 3D Gaussian Inverse Rendering for Relightable Scene Factorization (82 citations)
- Gamba: Marry Gaussian Splatting With Mamba for Single-View 3D Reconstruction (79 citations)
- HUGSIM: A Real-Time, Photo-Realistic and Closed-Loop Simulator for Autonomous Driving (68 citations)
- StylizedGS: Controllable Stylization for 3D Gaussian Splatting (61 citations)
- GaussNav: Gaussian Splatting for Visual Navigation (57 citations)
- DeferredGS: Decoupled and Relightable Gaussian Splatting With Deferred Shading (56 citations)
- Surfel-Based Gaussian Inverse Rendering for Fast and Relightable Dynamic Human Reconstruction From Monocular Videos (22 citations)
- GPS-Gaussian+: Generalizable Pixel-wise 3D Gaussian Splatting for Real-Time Human-Scene Rendering from Sparse Views (19 citations)
- Reconstructing Satellites in 3D From Amateur Telescope Images (13 citations)
- MCGS: Multiview Consistency Enhancement for Sparse-View 3D Gaussian Radiance Fields (12 citations)
and 10 more in the interactive map.
Face and shape · 22 papers
- TripoSG: High-Fidelity 3D Shape Synthesis Using Large-Scale Rectified Flow Models (225 citations)
- HHAvatar: Gaussian Head Avatar with Dynamic Hairs (43 citations)
- 3DCoMPaT ++ : An Improved Large-Scale 3D Vision Dataset for Compositional Recognition (15 citations)
- GPHM: Gaussian Parametric Head Model for Monocular Head Avatar Reconstruction (15 citations)
- SyncTalk++: High-Fidelity and Efficient Synchronized Talking Heads Synthesis Using Gaussian Splatting (14 citations)
- Spectrum-Enhanced Graph Attention Network for Garment Mesh Deformation (11 citations)
- MinD-3D++: Advancing fMRI-Based 3D Reconstruction With High-Quality Textured Mesh Generation and a Comprehensive Dataset (10 citations)
- GetMesh: A Controllable Model for High-quality Mesh Generation and Manipulation (8 citations)
- Towards High-Quality and Disentangled Face Editing in a 3D GAN (8 citations)
- Neural Vector Fields: Generalizing Distance Vector Fields by Codebooks and Zero-Curl Regularization (6 citations)
- Bridging Dimensions in Fingerprints to Advance Distinctiveness: Recovering 3D Minutiae From a Single Contactless 2D Fingerprint Image (4 citations)
- LBONet: Supervised Spectral Descriptors for Shape Analysis (4 citations)
and 10 more in the interactive map.
NeRF and radiance fields · 12 papers
- Explicit Correspondence Matching for Generalizable Neural Radiance Fields (61 citations)
- MoBluRF: Motion Deblurring Neural Radiance Fields for Blurry Monocular Video (19 citations)
- Geo-NI: Geometry-Aware Neural Interpolation for Light Field Rendering (15 citations)
- MS-NeRF: Multi-Space Neural Radiance Fields (9 citations)
- Dynamic Scene Understanding Through Object-Centric Voxelization and Neural Rendering (7 citations)
- Sparse-DeRF: Deblurred Neural Radiance Fields From Sparse View (6 citations)
- Dual-Camera All-in-Focus Neural Radiance Fields (5 citations)
- Learning Heterogeneous Mixture of Scene Experts for Large-Scale Neural Radiance Fields (4 citations)
- VD-NeRF: Visibility-Aware Decoupled Neural Radiance Fields for View-Consistent Editing and High-Frequency Relighting (4 citations)
- MAIR++: Improving Multi-View Attention Inverse Rendering With Implicit Lighting Representation (2 citations)
- SS-NeRF: Physically Based Sparse Spectral Rendering With Neural Radiance Field
- The NeRF Signature: Codebook-Aided Watermarking for Neural Radiance Fields
