mlsanityData to 13 Oct 2026

4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time

arXiv 2506.18890 ↗First posted 23 Jun 2025cs.CVcited by 13

Not on the map. Why?

Abstract (arXiv)

Can we scale 4D pretraining to learn general space-time representations that reconstruct an object from a few views at some times to any view at any time? We provide an affirmative answer with 4D-LRM, the first large-scale 4D reconstruction model that takes input from unconstrained views and timestamps and renders arbitrary novel view-time combinations. Unlike prior 4D approaches, e.g., optimization-based, geometry-based, or generative, that struggle with efficiency, generalization, or faithfulness, 4D-LRM learns a unified space-time representation and directly predicts per-pixel 4D Gaussian primitives from posed image tokens across time, enabling fast, high-quality rendering at, in principle, infinite frame rate. Our results demonstrate that scaling spatiotemporal pretraining enables accurate and efficient 4D reconstruction. We show that 4D-LRM generalizes to novel objects, interpolates across time, and handles diverse camera setups. It reconstructs 24-frame sequences in one forward pass with less than 1.5 seconds on a single A100 GPU.

Key facts. Cited by 13 papers in the map (data through 13 Oct 2026). Builds on 86 of its 100 references that are in the map. First posted on arXiv 23 Jun 2025.

Builds on

86 of 100 references.

References · 86

  1. 2017Attention Is All You Needcited by 60,106
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,950
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,043
  4. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,082
  5. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,170
  6. 2016Layer Normalizationcited by 5,624
  7. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,686
  8. 2023FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioningcited by 2,570
  9. 2020pixelNeRF: Neural Radiance Fields from One or Few Imagescited by 1,581
  10. 2023DUSt3R: Geometric 3D Vision Made Easycited by 1,539
  11. 2022Objaverse: A Universe of Annotated 3D Objectscited by 1,529
  12. 2025VGGT: Visual Geometry Grounded Transformercited by 1,497
38 more
  1. 2020D-NeRF: Neural Radiance Fields for Dynamic Scenescited by 1,462
  2. 2022TensoRF: Tensorial Radiance Fieldscited by 1,277
  3. 2021Efficient Geometry-aware 3D Generative Adversarial Networkscited by 1,205
  4. 2020Nerfies: Deformable Neural Radiance Fieldscited by 1,195
  5. 20234D Gaussian Splatting for Real-Time Dynamic Scene Renderingcited by 1,150
  6. 2017Mixed Precision Trainingcited by 1,147
  7. 20242D Gaussian Splatting for Geometrically Accurate Radiance Fieldscited by 1,070
  8. 2024Grounding Image Matching in 3D with MASt3Rcited by 965
  9. 2023LRM: Large Reconstruction Model for Single Image to 3Dcited by 843
  10. 2016Training Deep Nets with Sublinear Memory Costcited by 840
  11. 2023Objaverse-XL: A Universe of 10M+ 3D Objectscited by 771
  12. 2023K-Planes: Explicit Radiance Fields in Space, Time, and Appearancecited by 744
  13. 2021MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereocited by 738
  14. 2021Depth-supervised NeRF: Fewer Views and Faster Training for Freecited by 733
  15. 2024LGM: Large Multi-View Gaussian Model for High-Resolution 3D Content Creationcited by 732
  16. 2021IBRNet: Learning Multi-View Image-Based Renderingcited by 700
  17. 2023pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstructioncited by 673
  18. 2023Mip-Splatting: Alias-free 3D Gaussian Splattingcited by 663
  19. 2022Google Scanned Objects: A High-Quality Dataset of 3D Scanned Household Itemscited by 605
  20. 2023HexPlane: A Fast Representation for Dynamic Scenescited by 584
  21. 2025Continuous 3D Perception Model with Persistent Statecited by 534
  22. 2023Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splattingcited by 474
  23. 2020Non-Rigid Neural Radiance Fields: Reconstruction and Novel View Synthesis of a Dynamic Scene From Monocular Videocited by 471
  24. 2024MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motioncited by 437
  25. 2023Instant3D: Fast Text-to-3D with Sparse-View Generation and Large Reconstruction Modelcited by 376
  26. 2024SV3D: Novel Multi-view Synthesis and 3D Generation from a Single Image using Latent Video Diffusioncited by 364
  27. 2025Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Passcited by 358
  28. 2024GS-LRM: Large Reconstruction Model for 3D Gaussian Splattingcited by 345
  29. 2024Shape of Motion: 4D Reconstruction from a Single Videocited by 248
  30. 2017Photographic Image Synthesis with Cascaded Refinement Networkscited by 223
  31. 2023DMV3D: Denoising Multi-View Diffusion using 3D Large Reconstruction Modelcited by 213
  32. 2021GeoNeRF: Generalizing NeRF with Geometry Priorscited by 195
  33. 2023DreamGaussian4D: Generative 4D Gaussian Splattingcited by 193
  34. 2024LVSM: A Large View Synthesis Model with Minimal 3D Inductive Biascited by 182
  35. 2022SparseNeuS: Fast Generalizable Neural Surface Reconstruction from Sparse Viewscited by 176
  36. 2023PF-LRM: Pose-Free Large Reconstruction Model for Joint Pose and Shape Predictioncited by 165
  37. 2024CAT4D: Create Anything in 4D with Multi-View Video Diffusion Modelscited by 160
  38. 2022ARF: Artistic Radiance Fieldscited by 147

50 most cited of 86 All 86 (JSON)

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2025: 3; 2026: 10.

13 papers.

Sort:

By area:

  1. 2025Advances in Feed-Forward 3D Reconstruction and View Synthesis: A Surveycited by 25
  2. 2025UniLat3D: Geometry-Appearance Unified Latents for Single-Stage 3D Generationcited by 20
  3. 2026Motion 3-to-4: 3D Motion Reconstruction for 4D Synthesiscited by 15
  4. 2026Choreographing a World of Dynamic Objectscited by 5
  5. 2025Advances in 4D Representation: Geometry, Motion, and Interactioncited by 4
  6. 2026WildRayZer: Self-supervised Large View Synthesis in Dynamic Environmentscited by 4
  7. 2026Fast Spatial Memory with Elastic Test-Time Trainingcited by 3
  8. 2026LSRM: High-Fidelity Object-Centric Reconstruction via Scaled Context Windowscited by 2
  9. 2026Learning Global Motion with Compact Gaussians for Feed-Forward 4D Reconstructioncited by 2
  10. 2026HAT-4D: Lifting Monocular Video for 4D Multi-Object Interactions via Human-Agent Collaborationcited by 0
  11. 2026Seeing the World and the Self from Egocentric Videocited by 0
  12. 2026Eulerian Motion Reconstruction for Water Scenerycited by 0
1 more
  1. 2026Reconstructing the Dynamic World: A Representation-Centric View of 4D Scene Reconstructioncited by 0