mlsanityData to 7 Oct 2026

Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass

arXiv 2501.13928 ↗First posted 23 Jan 2025cs.CV cs.AI cs.GRcited by 356

Not on the map. Why?

Abstract (arXiv)

Multi-view 3D reconstruction remains a core challenge in computer vision, particularly in applications requiring accurate and scalable representations across diverse perspectives. Current leading methods such as DUSt3R employ a fundamentally pairwise approach, processing images in pairs and necessitating costly global alignment procedures to reconstruct from multiple views. In this work, we propose Fast 3D Reconstruction (Fast3R), a novel multi-view generalization to DUSt3R that achieves efficient and scalable 3D reconstruction by processing many views in parallel. Fast3R's Transformer-based architecture forwards N images in a single forward pass, bypassing the need for iterative alignment. Through extensive experiments on camera pose estimation and 3D reconstruction, Fast3R demonstrates state-of-the-art performance, with significant improvements in inference speed and reduced error accumulation. These results establish Fast3R as a robust alternative for multi-view applications, offering enhanced scalability without compromising reconstruction accuracy.

Key facts. Cited by 356 papers in the map (data through 7 Oct 2026). Builds on 55 of its 74 references that are in the map. First posted on arXiv 23 Jan 2025.

Builds on

55 of 74 references.

References · 55

  1. 2017Attention Is All You Needcited by 59,932
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  4. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,611
  5. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  6. 2023FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioningcited by 2,551
  7. 2020Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domainscited by 2,170
  8. 2019Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelismcited by 1,829
  9. 2021Vision Transformers for Dense Predictioncited by 1,789
  10. 2017Decoupled Weight Decay Regularizationcited by 1,722
  11. 2023DUSt3R: Geometric 3D Vision Made Easycited by 1,526
  12. 2019Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfercited by 1,484
38 more
  1. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,437
  2. 2024Depth Anything: Unleashing the Power of Large-Scale Unlabeled Datacited by 1,346
  3. 2017SuperPoint: Self-Supervised Interest Point Detection and Descriptioncited by 1,313
  4. 2019ZeRO: Memory Optimizations Toward Training Trillion Parameter Modelscited by 1,278
  5. 2019Habitat: A Platform for Embodied AI Researchcited by 1,163
  6. 2015ORB-SLAM: a Versatile and Accurate Monocular SLAM Systemcited by 983
  7. 2024Grounding Image Matching in 3D with MASt3Rcited by 960
  8. 2016ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D Camerascited by 801
  9. 2023ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenescited by 747
  10. 2018MegaDepth: Learning Single-View Depth Prediction from Internet Photoscited by 731
  11. 2023Extending Context Window of Large Language Models via Positional Interpolationcited by 721
  12. 2021DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Camerascited by 673
  13. 2021Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstructioncited by 646
  14. 2020Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understandingcited by 593
  15. 2020TartanAir: A Dataset to Push the Limits of Visual SLAMcited by 494
  16. 2019BlendedMVS: A Large-scale Dataset for Generalized Multi-view Stereo Networkscited by 487
  17. 2023Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectivescited by 477
  18. 2024MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motioncited by 431
  19. 2021Omnidata: A Scalable Pipeline for Making Multi-Task Mid-Level Vision Datasets from 3D Scanscited by 336
  20. 2019D2-Net: A Trainable CNN for Joint Detection and Description of Local Featurescited by 324
  21. 2021Neural RGB-D Surface Reconstructioncited by 312
  22. 2016LIFT: Learned Invariant Feature Transformcited by 305
  23. 2023PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point Trackingcited by 274
  24. 20243D Reconstruction with Spatial Memorycited by 269
  25. 2022TAP-Vid: A Benchmark for Tracking Any Point in a Videocited by 258
  26. 2023Visual Geometry Grounded Deep Structure From Motioncited by 225
  27. 2018Stereo Magnification: Learning View Synthesis using Multiplane Imagescited by 206
  28. 2021Pixel-Perfect Structure-from-Motion with Featuremetric Refinementcited by 199
  29. 2018Mesh-TensorFlow: Deep Learning for Supercomputerscited by 168
  30. 2022CroCo: Self-Supervised Pre-training for 3D Vision Tasks by Cross-View Completioncited by 158
  31. 2023PoseDiffusion: Solving Pose Estimation via Diffusion-aided Bundle Adjustmentcited by 146
  32. 2023Infinite Photorealistic Worlds using Procedural Generationcited by 142
  33. 2024Cameras as Rays: Pose Estimation via Ray Diffusioncited by 129
  34. 2023PoSE: Efficient Context Window Extension of LLMs via Positional Skip-wise Trainingcited by 109
  35. 2022RelPose: Predicting Probabilistic Relative Rotation for Single Objects in the Wildcited by 102
  36. 2022ParticleSfM: Exploiting Dense Point Trajectories for Localizing Moving Cameras in the Wildcited by 90
  37. 2020PyTorch Distributed: Experiences on Accelerating Data Parallel Trainingcited by 87
  38. 2023RelPose++: Recovering 6D Poses from Sparse-view Observationscited by 87

50 most cited of 55 All 55 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 148; 2026: 208.

356 papers.

Sort:

By area:

  1. 2025VGGT: Visual Geometry Grounded Transformercited by 1,475
  2. 2025Depth Anything 3: Recovering the Visual Space from Any Viewscited by 626
  3. 2025MapAnything: Universal Feed-Forward Metric 3D Reconstructioncited by 295
  4. 2025π3: Permutation-Equivariant Visual Geometry Learningcited by 292
  5. 2025AnySplat: Feed-forward 3D Gaussian Splatting from Unconstrained Viewscited by 234
  6. 2025ViPE: Video Pose Engine for 3D Geometric Perceptioncited by 162
  7. 2025VLM-3R: Vision-Language Models Augmented with Instruction-Aligned 3D Reconstructioncited by 154
  8. 2025Streaming 4D Visual Geometry Transformercited by 147
  9. 2025TTT3R: 3D Reconstruction as Test-Time Trainingcited by 122
  10. 2025VGGT-Long: Chunk it, Loop it, Align it -- Pushing VGGT's Limits on Kilometer-scale Long RGB Sequencescited by 107
  11. 2025FastVGGT: Training-Free Acceleration of Visual Geometry Transformercited by 93
  12. 2025Geometry Forcing: Marrying Video Diffusion and 3D Representation for Consistent World Modelingcited by 86
38 more
  1. 2025Point3R: Streaming 3D Reconstruction with Explicit Spatial Pointer Memorycited by 82
  2. 2025JanusVLN: Decoupling Semantics and Spatiality with Dual Implicit Memory for Vision-Language Navigationcited by 80
  3. 2025STream3R: Scalable Sequential 3D Reconstruction with Causal Transformercited by 76
  4. 2025RayZer: A Self-supervised Large View Synthesis Modelcited by 74
  5. 2025SpatialVID: A Large-Scale Video Dataset with Spatial Annotationscited by 72
  6. 2025Geo4D: Leveraging Video Generators for Geometric 4D Scene Reconstructioncited by 66
  7. 2025WorldMirror: Universal 3D World Reconstruction with Any-Prior Promptingcited by 60
  8. 2025OmniWorld: A Multi-Domain and Multi-Modal Dataset for 4D World Modelingcited by 59
  9. 2025Multi-SpatialMLLM: Multi-Frame Spatial Understanding with Multi-Modal Large Language Modelscited by 52
  10. 2026InfiniteVGGT: Visual Geometry Grounded Transformer for Endless Streamscited by 51
  11. 2025Memory Forcing: Spatio-Temporal Memory for Consistent Scene Generation on Minecraftcited by 46
  12. 2025Efficiently Reconstructing Dynamic Scenes One D4RT at a Timecited by 46
  13. 2025LONG3R: Long Sequence Streaming 3D Reconstructioncited by 45
  14. 2025Dens3R: A Foundation Model for 3D Geometry Predictioncited by 44
  15. 2025G2VLM: Geometry Grounded Vision Language Model with Unified 3D Reconstruction and Spatial Reasoningcited by 39
  16. 2025Human3R: Everyone Everywhere All at Oncecited by 35
  17. 2025YoNoSplat: You Only Need One Model for Feedforward 3D Gaussian Splattingcited by 34
  18. 2026HY-World 2.0: A Multi-Modal World Model for Reconstructing, Generating, and Simulating 3D Worldscited by 34
  19. 2025FantasyWorld: Geometry-Consistent World Modeling via Unified Video and 3D Predictioncited by 33
  20. 2025Spatia: Video Generation with Updatable Spatial Memorycited by 33
  21. 2025WinT3R: Window-Based Streaming Reconstruction with Camera Token Poolcited by 32
  22. 2026LingBot-Map: Geometric Context Transformer for Streaming 3D Reconstructioncited by 32
  23. 2025MoVieS: Motion-Aware 4D Dynamic View Synthesis in One Secondcited by 31
  24. 2025Trace Anything: Representing Any Video in 4D via Trajectory Fieldscited by 31
  25. 2025What Is The Best 3D Scene Representation for Robotics? From Geometric to Foundation Modelscited by 29
  26. 2026V-DPM: 4D Video Reconstruction with Dynamic Point Mapscited by 29
  27. 2025FlowR: Flowing from Sparse to Dense 3D Reconstructionscited by 28
  28. 2025VGGT4D: Mining Motion Cues in Visual Geometry Transformers for 4D Scene Reconstructioncited by 26
  29. 2026Scal3R: Scalable Test-Time Training for Large-Scale 3D Reconstructioncited by 25
  30. 2025Geometry-aware 4D Video Generation for Robot Manipulationcited by 23
  31. 2025Advances in Feed-Forward 3D Reconstruction and View Synthesis: A Surveycited by 23
  32. 2025LongSplat: Robust Unposed 3D Gaussian Splatting for Casual Long Videoscited by 23
  33. 2026ZipMap: Linear-Time Stateful 3D Reconstruction via Test-Time Trainingcited by 23
  34. 2025Block-Sparse Global Attention for Efficient Multi-View Geometry Transformerscited by 22
  35. 2025UniLat3D: Geometry-Appearance Unified Latents for Single-Stage 3D Generationcited by 20
  36. 2026VGGT-Omegacited by 20
  37. 2025Test3R: Learning to Reconstruct 3D at Test Timecited by 19
  38. 2025Viser: Imperative, Web-based 3D Visualization in Pythoncited by 19

50 most cited of 356 All 356 (JSON)