mlsanityData to 7 Oct 2026

D2USt3R: Enhancing 3D Reconstruction for Dynamic Scenes

arXiv 2504.06264 ↗First posted 8 Apr 2025cs.CVcited by 10

Not on the map. Why?

Abstract (arXiv)

In this work, we address the task of 3D reconstruction in dynamic scenes, where object motions frequently degrade the quality of previous 3D pointmap regression methods, such as DUSt3R, that are originally designed for static 3D scene reconstruction. Although these methods provide an elegant and powerful solution in static settings, they struggle in the presence of dynamic motions that disrupt alignment based solely on camera poses. To overcome this, we propose D2USt3R that directly regresses Static-Dynamic Aligned Pointmaps (SDAP) that simultaneiously capture both static and dynamic 3D scene geometry. By explicitly incorporating both spatial and temporal aspects, our approach successfully encapsulates 3D dense correspondence to the proposed pointmaps, enhancing downstream tasks. Extensive experimental evaluations demonstrate that our proposed approach consistently achieves superior 3D reconstruction performance across various datasets featuring complex motions.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 40 of its 57 references that are in the map. First posted on arXiv 8 Apr 2025.

Builds on

40 of 57 references.

References · 40

  1. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,141
  2. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  3. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,385
  4. 2021Vision Transformers for Dense Predictioncited by 1,789
  5. 2023DUSt3R: Geometric 3D Vision Made Easycited by 1,526
  6. 2024Depth Anything V2cited by 1,385
  7. 2024Grounding Image Matching in 3D with MASt3Rcited by 960
  8. 2017The 2017 DAVIS Challenge on Video Object Segmentationcited by 942
  9. 2020ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAMcited by 862
  10. 2018Pixel2Mesh: Generating 3D Mesh Models from Single RGB Imagescited by 649
  11. 20163D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstructioncited by 587
  12. 2019Deep Learning for 3D Point Clouds: A Surveycited by 557
28 more
  1. 2025Continuous 3D Perception Model with Persistent Statecited by 527
  2. 2020TartanAir: A Dataset to Push the Limits of Visual SLAMcited by 494
  3. 2024MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motioncited by 431
  4. 2018DeepVoxels: Learning Persistent 3D Feature Embeddingscited by 401
  5. 2025Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Passcited by 356
  6. 2018GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera Posecited by 341
  7. 2023Repurposing Diffusion-Based Image Generators for Monocular Depth Estimationcited by 323
  8. 2023PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point Trackingcited by 274
  9. 2023Visual Geometry Grounded Deep Structure From Motioncited by 225
  10. 2024MegaSaM: Accurate, Fast, and Robust Structure and Motion from Casual Dynamic Videoscited by 224
  11. 2021Pixel-Perfect Structure-from-Motion with Featuremetric Refinementcited by 199
  12. 2019ReFusion: 3D Reconstruction in Dynamic Environments for RGB-D Cameras Exploiting Residualscited by 180
  13. 2023Spring: A High-Resolution High-Detail Dataset and Benchmark for Scene Flow, Optical Flow and Stereocited by 170
  14. 2024SEA-RAFT: Simple, Efficient, Accurate RAFT for Optical Flowcited by 169
  15. 2022CroCo v2: Improved Cross-view Completion Pre-training for Stereo Matching and Optical Flowcited by 165
  16. 2017Learning Efficient Point Cloud Generation for Dense 3D Object Reconstructioncited by 133
  17. 2022Cost Aggregation with 4D Convolutional Swin Transformer for Few-Shot Segmentationcited by 114
  18. 2022ParticleSfM: Exploiting Dense Point Trajectories for Localizing Moving Cameras in the Wildcited by 90
  19. 2024Align3R: Aligned Monocular Depth Estimation for Dynamic Videoscited by 79
  20. 2021CATs: Cost Aggregation Transformers for Visual Correspondencecited by 77
  21. 2024PF3plat: Pose-Free Feed-Forward 3D Gaussian Splattingcited by 60
  22. 2022CATs++: Boosting Cost Aggregation with Convolutions and Transformerscited by 53
  23. 2023Unifying Correspondence, Pose and NeRF for Pose-Free Novel View Synthesis from Stereo Pairscited by 37
  24. 2022Neural Matching Fields: Implicit Representation of Matching Fields for Visual Correspondencecited by 32
  25. 2021Deep Matching Prior: Test-Time Optimization for Dense Correspondencecited by 30
  26. 2024Cross-View Completion Models are Zero-shot Correspondence Estimatorscited by 30
  27. 2025Zero-Shot Monocular Scene Flow Estimation in the Wildcited by 20
  28. 2024BlinkVision: A Benchmark for Optical Flow, Scene Flow and Point Tracking Estimation using RGB Frames and Eventscited by 13

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 6; 2026: 4.

10 papers.

  1. 2025RealX3D: A Physically-Degraded 3D Benchmark for Multi-view Visual Restoration and Reconstructioncited by 19
  2. 2025Reconstructing 4D Spatial Intelligence: A Surveycited by 13
  3. 2026Flow4R: Unifying 4D Reconstruction and Tracking with Scene Flowcited by 8
  4. 2025C4D: 4D Made from 3D through Dual Correspondencescited by 7
  5. 2025E3D-Bench: A Benchmark for End-to-End 3D Geometric Foundation Modelscited by 6
  6. 2025LASER: Layer-wise Scale Alignment for Training-Free Streaming 4D Reconstructioncited by 5
  7. 2026TrackCraft3R: Repurposing Video Diffusion Transformers for Dense 3D Trackingcited by 3
  8. 2025ProDyG: Progressive Dynamic Scene Reconstruction via Gaussian Splatting from Monocular Videoscited by 1
  9. 2026TurboVGGT: Fast Visual Geometry Reconstruction with Adaptive Alternating Attentioncited by 1
  10. 2026UniQuery4R: Unified 4D Scene Reconstruction from a Single Querycited by 1