mlsanityData to 14 Oct 2026

Advances in Feed-Forward 3D Reconstruction and View Synthesis: A Survey

arXiv 2507.14501 ↗First posted 19 Jul 2025cs.CVcited by 27

Not on the map. Why?

Abstract (arXiv)

3D reconstruction and view synthesis are foundational problems in computer vision, graphics, and immersive technologies such as augmented reality (AR), virtual reality (VR), and digital twins. Traditional methods rely on computationally intensive iterative optimization in a complex chain, limiting their applicability in real-world scenarios. Recent advances in feed-forward approaches, driven by deep learning, have revolutionized this field by enabling fast and generalizable 3D reconstruction and view synthesis. This survey offers a comprehensive review of feed-forward techniques for 3D reconstruction and view synthesis, with a taxonomy according to the underlying representation architectures including point cloud, 3D Gaussian Splatting (3DGS), Neural Radiance Fields (NeRF), etc. We examine key tasks such as pose-free reconstruction, dynamic 3D reconstruction, and 3D-aware image and video synthesis, highlighting their applications in digital humans, SLAM, robotics, and beyond. In addition, we review commonly used datasets with detailed statistics, along with evaluation protocols for various downstream tasks. We conclude by discussing open research challenges and promising directions for future work, emphasizing the potential of feed-forward approaches to advance the state of the art in 3D vision.

Key facts. Cited by 27 papers in the map (data through 14 Oct 2026). Builds on 278 of its 307 references that are in the map. First posted on arXiv 19 Jul 2025.

Builds on

278 of 307 references.

References · 278

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  4. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  5. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  6. 2017Mask R-CNNcited by 7,277
  7. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,226
  8. 2022Scalable Diffusion Models with Transformerscited by 6,054
  9. 2017Neural Discrete Representation Learningcited by 4,712
  10. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,356
  11. 2020Taming Transformers for High-Resolution Image Synthesiscited by 3,253
  12. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
38 more
  1. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
  2. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,689
  3. 2023FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioningcited by 2,598
  4. 2022Video Diffusion Modelscited by 2,305
  5. 2024CogVideoX: Text-to-Video Diffusion Models with An Expert Transformercited by 2,261
  6. 2018Occupancy Networks: Learning 3D Reconstruction in Function Spacecited by 2,041
  7. 2021Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fieldscited by 2,032
  8. 2021Vision Transformers for Dense Predictioncited by 1,803
  9. 2020pixelNeRF: Neural Radiance Fields from One or Few Imagescited by 1,584
  10. 2023DUSt3R: Geometric 3D Vision Made Easycited by 1,567
  11. 2022Objaverse: A Universe of Annotated 3D Objectscited by 1,540
  12. 2025VGGT: Visual Geometry Grounded Transformercited by 1,539
  13. 2023Zero-1-to-3: Zero-shot One Image to 3D Objectcited by 1,489
  14. 2017Matterport3D: Learning from RGB-D Data in Indoor Environmentscited by 1,376
  15. 2024Depth Anything: Unleashing the Power of Large-Scale Unlabeled Datacited by 1,360
  16. 2024Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Dualitycited by 1,336
  17. 2021Efficient Geometry-aware 3D Generative Adversarial Networkscited by 1,205
  18. 2024Grounding Image Matching in 3D with MASt3Rcited by 983
  19. 2016A Point Set Generation Network for 3D Object Reconstruction from a Single Imagecited by 979
  20. 2023MVDream: Multi-view Diffusion for 3D Generationcited by 938
  21. 2019The Replica Dataset: A Digital Replica of Indoor Spacescited by 904
  22. 2023LRM: Large Reconstruction Model for Single Image to 3Dcited by 849
  23. 2018MVSNet: Depth Inference for Unstructured Multi-view Stereocited by 788
  24. 2020GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fieldscited by 785
  25. 2023Objaverse-XL: A Universe of 10M+ 3D Objectscited by 774
  26. 2023ScanNet++: A High-Fidelity Dataset of 3D Indoor Scenescited by 764
  27. 2020Convolutional Occupancy Networkscited by 761
  28. 2021MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereocited by 740
  29. 2024LGM: Large Multi-View Gaussian Model for High-Resolution 3D Content Creationcited by 739
  30. 2018MegaDepth: Learning Single-View Depth Prediction from Internet Photoscited by 737
  31. 2020pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesiscited by 708
  32. 2021IBRNet: Learning Multi-View Image-Based Renderingcited by 703
  33. 2022Point-E: A System for Generating 3D Point Clouds from Complex Promptscited by 699
  34. 2020GRAF: Generative Radiance Fields for 3D-Aware Image Synthesiscited by 689
  35. 2023Wonder3D: Single Image to 3D using Cross-Domain Diffusioncited by 689
  36. 2021DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Camerascited by 687
  37. 2023pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstructioncited by 684
  38. 2021Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstructioncited by 652

50 most cited of 278 All 278 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 6; 2026: 21.

27 papers.

Sort:

By area:

  1. 2025TTT3R: 3D Reconstruction as Test-Time Trainingcited by 129
  2. 2025PAGE-4D: Disentangled pose and geometry estimation for vggt-4d perceptioncited by 8
  3. 2026Feed-Forward 3D Scene Modeling: A Problem-Driven Perspectivecited by 6
  4. 2025S2-MLLM: Boosting Spatial Reasoning Capability of MLLMs for 3D Visual Grounding with Structural Guidancecited by 5
  5. 2026Sat3DGen: Comprehensive Street-Level 3D Scene Generation from Single Satellite Imagecited by 5
  6. 2026Speeding Up the Learning of 3D Gaussians with Much Shorter Gaussian Listscited by 4
  7. 2026PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Spacecited by 4
  8. 2026TriSplat: Simulation-Ready Feed-Forward 3D Scene Reconstructioncited by 3
  9. 2025MuSASplat: Efficient Sparse-View 3D Gaussian Splats via Lightweight Multi-Scale Adaptationcited by 2
  10. 2026OneWorld: Taming Scene Generation with 3D Unified Representation Autoencodercited by 2
  11. 2026VersaQ-3D: Architecture Support for Visual Geometry Grounded Transformers via Versatile Quantizationcited by 1
  12. 2026KitchenTwin: Semantically and Geometrically Grounded 3D Kitchen Digital Twinscited by 1
15 more
  1. 2026Uncertainty Quality of VGGT: An Analysis on the DTU Benchmark Datasetcited by 1
  2. 2026Clarify Before Executing: A Self-Evolving Agent for Resolving Intent Asymmetry in 3D Tool Orchestrationcited by 1
  3. 2026Think Locally, Refine Globally for Memory-Efficient 3D Reconstructioncited by 1
  4. 2026Reconstructing the Dynamic World: A Representation-Centric View of 4D Scene Reconstructioncited by 1
  5. 2025Room Envelopes: A Synthetic Dataset for Indoor Layout Reconstruction from Imagescited by 0
  6. 2025Make-It-Poseable: Feed-forward Latent Posing Model for 3D Characterscited by 0
  7. 2026ReCoSplat: Online Feed-Forward Gaussian Splatting via Render-and-Comparecited by 0
  8. 2026DriveFix: Spatio-Temporally Coherent Driving Scene Restorationcited by 0
  9. 2026AdvSplat: Adversarial Attacks on Feed-Forward Gaussian Splatting Modelscited by 0
  10. 2026FurnSet: Exploiting Repeats for 3D Scene Reconstructioncited by 0
  11. 2026EigeNet: Geometry-Informed Multi-Modal Learning for Few-shot Novel View RIR Predictioncited by 0
  12. 2026Effective Multi-sensor Conditioning for Street-view Novel-view Synthesiscited by 0
  13. 2026CasDeblurGS: Cascaded 2D-to-3D Multi-View Consistency for 3D Gaussian Splatting from Two Blurry Imagescited by 0
  14. 2026LightBridge: Feed-Forward Generative Relighting for 3D Gaussian Splattingcited by 0
  15. 2026CoTL: Compact Token Learning for 3D Scene Representationcited by 0