ICLR 2025 accepted papers on 3D vision
133 of the 3,703 papers accepted at ICLR 2025 (3.6%) are on 3D vision, compared with 3.3% at ICLR 2024. They fall into 6 topics.
See ICLR 2025 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
Gaussian splatting · 45 papers
- On Scaling Up 3D Gaussian Splatting Training (19 citations)
- CityGaussianV2: Efficient and Geometrically Accurate Reconstruction for Large-Scale Scenes (3 citations)
- Locality-aware Gaussian Compression for Fast and High-quality Rendering (3 citations)
- Dynamic Gaussians Mesh: Consistent Mesh Reconstruction from Dynamic Scenes (2 citations)
- HiSplat: Hierarchical 3D Gaussian Splatting for Generalizable Sparse-View Reconstruction (2 citations)
- OMG: Opacity Matters in Material Modeling with Gaussian Splatting (2 citations)
- 3DitScene: Editing Any Scene via Language-guided Disentangled Gaussian Splatting (1 citation)
- Fast Feedforward 3D Gaussian Splatting Compression (1 citation)
- GI-GS: Global Illumination Decomposition on Gaussian Splatting for Inverse Rendering (1 citation)
- GS-CPR: Efficient Camera Pose Refinement via 3D Gaussian Splatting (1 citation)
- OmniRe: Omni Urban Scene Reconstruction (1 citation)
- STORM: Spatio-TempOral Reconstruction Model For Large-Scale Outdoor Scenes (1 citation)
and 33 more in the interactive map.
3D generation · 42 papers
- IPDreamer: Appearance-Controllable 3D Object Generation with Complex Image Prompts (4 citations)
- Generating CAD Code with Vision-Language Models for 3D Designs (3 citations)
- Controllable Satellite-to-Street-View Synthesis with Precise Pose Alignment and Zero-Shot Environmental Control (1 citation)
- CubeDiff: Repurposing Diffusion-Based Image Models for Panorama Generation (1 citation)
- Diff3DS: Generating View-Consistent 3D Sketch via Differentiable Curve Rendering (1 citation)
- DiffSplat: Repurposing Image Diffusion Models for Scalable Gaussian Splat Generation (1 citation)
- Diffusion$^2$: Dynamic 3D Content Generation via Score Composition of Video and Multi-view Diffusion Models (1 citation)
- EmbodiedSAM: Online Segment Any 3D Thing in Real Time (1 citation)
- GenXD: Generating Any 3D and 4D Scenes (1 citation)
- Layout-your-3D: Controllable and Precise 3D Generation with 2D Blueprint (1 citation)
- Learning General-purpose Biomedical Volume Representations using Randomized Synthesis (1 citation)
- A3D: Does Diffusion Dream about 3D Alignment?
and 30 more in the interactive map.
Face and shape · 23 papers
- MeshAnything: Artist-Created Mesh Generation with Autoregressive Transformers (8 citations)
- TetSphere Splatting: Representing High-Quality Geometry with Lagrangian Volumetric Meshes (2 citations)
- DICE: End-to-end Deformation Capture of Hand-Face Interactions from a Single Image (1 citation)
- PivotMesh: Generic 3D Mesh Generation via Pivot Vertices Guidance (1 citation)
- Real2Code: Reconstruct Articulated Objects via Code Generation (1 citation)
- SurFhead: Affine Rig Blending for Geometrically Accurate 2D Gaussian Surfel Head Avatars (1 citation)
- TEASER: Token Enhanced Spatial Modeling for Expressions Reconstruction (1 citation)
- Differentiable and Learnable Wireless Simulation with Geometric Transformers
- FreSh: Frequency Shifting for Accelerated Neural Representation Learning
- GALA: Geometry-Aware Local Adaptive Grids for Detailed 3D Generation
- Gaussian Head & Shoulders: High Fidelity Neural Upper Body Avatars with Anchor Gaussian Guided Texture Warping
- Generalizable Human Gaussians from Single-View Image
and 11 more in the interactive map.
Depth and stereo · 11 papers
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (20 citations)
- MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion (3 citations)
- RESfM: Robust Deep Equivariant Structure from Motion (1 citation)
- DELTA: DENSE EFFICIENT LONG-RANGE 3D TRACKING FOR ANY VIDEO
- LoRA3D: Low-Rank Self-Calibration of 3D Geometric Foundation models
- Neural Eulerian Scene Flow Fields
- Optimizing 4D Gaussians for Dynamic Scene Video from Single Landscape Images
- Self-supervised Monocular Depth Estimation Robust to Reflective Surface Leveraged by Triplet Mining
- Unposed Sparse Views Room Layout Reconstruction in the Age of Pretrain Model
- Controllable Blur Data Augmentation Using 3D-Aware Motion Estimation
- Revealing the 3D Cosmic Web through Gravitationally Constrained Neural Fields
Human pose estimation · 8 papers
- ImDy: Human Inverse Dynamics from Imitated Observations (1 citation)
- 6D Object Pose Tracking in Internet Videos for Robotic Manipulation
- CapeX: Category-Agnostic Pose Estimation from Textual Point Explanation
- CHAMP: Conformalized 3D Human Multi-Hypothesis Pose Estimators
- CoMotion: Concurrent Multi-person 3D Motion
- Learning Shape-Independent Transformation via Spherical Representations for Category-Level Object Pose Estimation
- SiMHand: Mining Similar Hands for Large-Scale 3D Hand Pose Pre-training
- What's the Move? Hybrid Imitation Learning via Salient Points