mlsanityData to 7 Oct 2026

L3DG: Latent 3D Gaussian Diffusion

arXiv 2410.13530 ↗First posted 17 Oct 2024cs.CV cs.GRcited by 42

Not on the map. Why?

Abstract (arXiv)

We propose L3DG, the first approach for generative 3D modeling of 3D Gaussians through a latent 3D Gaussian diffusion formulation. This enables effective generative 3D modeling, scaling to generation of entire room-scale scenes which can be very efficiently rendered. To enable effective synthesis of 3D Gaussians, we propose a latent diffusion formulation, operating in a compressed latent space of 3D Gaussians. This compressed latent space is learned by a vector-quantized variational autoencoder (VQ-VAE), for which we employ a sparse convolutional architecture to efficiently operate on room-scale scenes. This way, the complexity of the costly generation process via diffusion is substantially reduced, allowing higher detail on object-level generation, as well as scalability to large scenes. By leveraging the 3D Gaussian representation, the generated scenes can be rendered from arbitrary viewpoints in real-time. We demonstrate that our approach significantly improves visual quality over prior work on unconditional object-level radiance field synthesis and showcase its applicability to room-scale scene generation.

Key facts. Cited by 42 papers in the map (data through 7 Oct 2026). Builds on 34 of its 52 references that are in the map. First posted on arXiv 17 Oct 2024.

Builds on

34 of 52 references.

References · 34

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  4. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
  5. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,141
  6. 2022Generative Adversarial Networkscited by 5,603
  7. 2017Neural Discrete Representation Learningcited by 4,650
  8. 2022DreamFusion: Text-to-3D using 2D Diffusioncited by 3,006
  9. 2018Demystifying MMD GANscited by 1,268
  10. 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,201
  11. 2020GIRAFFE: Representing Scenes as Compositional Generative Neural Feature Fieldscited by 784
  12. 2021Diffusion Probabilistic Models for 3D Point Cloud Generationcited by 734
22 more
  1. 2020pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesiscited by 708
  2. 2020GRAF: Generative Radiance Fields for 3D-Aware Image Synthesiscited by 688
  3. 2023SyncDreamer: Generating Multiview-consistent Images from a Single-view Imagecited by 638
  4. 2022LION: Latent Point Diffusion Models for 3D Shape Generationcited by 552
  5. 2021StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesiscited by 477
  6. 2022GET3D: A Generative Model of High Quality 3D Textured Shapes Learned from Imagescited by 446
  7. 20203D-FRONT: 3D Furnished Rooms with layOuts and semaNTicscited by 375
  8. 2021ABO: Dataset and Benchmarks for Real-World 3D Object Understandingcited by 360
  9. 2022Novel View Synthesis with Diffusion Modelscited by 299
  10. 2023Text-to-3D using Gaussian Splattingcited by 299
  11. 2023GaussianDreamer: Fast Generation from Text to 3D Gaussians by Bridging 2D and 3D Diffusion Modelscited by 255
  12. 2020Learning Gradient Fields for Shape Generationcited by 245
  13. 2022DiffRF: Rendering-Guided 3D Radiance Field Diffusioncited by 204
  14. 2017Octree Generating Networks: Efficient Convolutional Architectures for High-resolution 3D Outputscited by 189
  15. 2023Single-Stage Diffusion NeRF: A Unified Approach to 3D Generation and Reconstructioncited by 180
  16. 2022VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Gridscited by 133
  17. 2024latentSplat: Autoencoding Variational Gaussians for Fast Generalizable 3D Reconstructioncited by 114
  18. 2022EpiGRAF: Rethinking training of 3D GANscited by 108
  19. 2019Towards Unsupervised Learning of Generative Models for 3D Controllable Image Synthesiscited by 104
  20. 2020Generative Sparse Detection Networks for 3D Single-shot Object Detectioncited by 69
  21. 2023GaussianDiffusion: 3D Gaussian Splatting for Denoising Diffusion Probabilistic Models with Structured Noisecited by 30
  22. 2018PhotoShape: Photorealistic Materials for Large-Scale Shape Collectionscited by 21

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 7; 2025: 18; 2026: 17.

42 papers.

Sort:

By area:

  1. 2024MIDI: Multi-Instance Diffusion for Single Image to 3D Scene Generationcited by 113
  2. 2025Bolt3D: Generating 3D Scenes in Secondscited by 57
  3. 2025Amodal3R: Amodal 3D Reconstruction from Occluded 2D Imagescited by 52
  4. 2024MeshArt: Generating Articulated Meshes with Structure-Guided Transformerscited by 40
  5. 2025VoxHammer: Training-Free Precise and Coherent 3D Editing in Native 3D Spacecited by 40
  6. 2025Efficient Diffusion Models: A Surveycited by 35
  7. 2024Prometheus: 3D-Aware Latent Diffusion Models for Feed-Forward Text-to-3D Scene Generationcited by 33
  8. 2025OctGPT: Octree-based Multiscale Autoregressive Models for 3D Shape Generationcited by 27
  9. 2026Gen3R: 3D Scene Generation Meets Feed-Forward Reconstructioncited by 24
  10. 2024HybridGS: Decoupling Transients and Statics with 2D and 3D Gaussian Splattingcited by 16
  11. 2025SceneSplat++: A Large Dataset and Comprehensive Benchmark for Language Gaussian Splattingcited by 14
  12. 2025QuickSplat: Fast 3D Surface Reconstruction via Learned Gaussian Initializationcited by 11
30 more
  1. 2024A Lesson in Splats: Teacher-Guided Diffusion for 3D Gaussian Splats Generation with 2D Supervisioncited by 9
  2. 2026SegviGen: Repurposing 3D Generative Model for Part Segmentationcited by 9
  3. 2024DNF: Unconditional 4D Generation with Dictionary-based Neural Fieldscited by 8
  4. 2025Repurposing 2D Diffusion Models with Gaussian Atlas for 3D Generationcited by 7
  5. 2025Progressive Rendering Distillation: Adapting Stable Diffusion for Instant Text-to-Mesh Generation without 3D Datacited by 7
  6. 2025SPARK: Sim-ready Part-level Articulated Reconstruction with VLM Knowledgecited by 6
  7. 2026SceneTok: A Compressed, Diffusable Token Space for 3D Scenescited by 6
  8. 2026WorldClaw: Agentic 3D Open-World Generation at Scalecited by 6
  9. 2025Can3Tok: Canonical 3D Tokenization and Latent Modeling of Scene-Level 3D Gaussianscited by 5
  10. 2026GaussianGPT: Towards Autoregressive 3D Gaussian Scene Generationcited by 4
  11. 2026GenAssets: Generating in-the-wild 3D Assets in Latent Spacecited by 4
  12. 2025GSV3D: Gaussian Splatting-based Geometric Distillation with Stable Video Diffusion for Single-Image 3D Object Generationcited by 3
  13. 2026Repurposing 3D Generative Model for Autoregressive Layout Generationcited by 3
  14. 2026Free-Range Gaussians: Non-Grid-Aligned Generative 3D Gaussian Reconstructioncited by 2
  15. 2026Beyond Voxel 3D Editing: Learning from 3D Masks and Self-Constructed Datacited by 2
  16. 2025HuGDiffusion: Generalizable Single-Image Human Rendering via 3D Gaussian Diffusioncited by 1
  17. 2025Object-X: Learning to Reconstruct Multi-Modal 3D Object Representationscited by 1
  18. 2025Hyper Diffusion Avatars: Dynamic Human Avatar Generation using Network Weight Space Diffusioncited by 1
  19. 2025From Fields to Splats: A Cross-Domain Survey of Real-Time Neural Scene Representationscited by 1
  20. 2026ScenDi: 3D-to-2D Scene Diffusion Cascades for Urban Generationcited by 1
  21. 2026BoxSplitGen: A Generative Model for 3D Part Bounding Boxes in Varying Granularitycited by 1
  22. 2026OrbitNVS: Harnessing Video Diffusion Priors for Novel View Synthesiscited by 1
  23. 2024Geometry Distributionscited by 0
  24. 2025Squeeze3D: Extreme Neural Compression with Latent Space Bridgingcited by 0
  25. 2025CC-FMO: Camera-Conditioned Zero-Shot Single Image to 3D Scene Generation with Foundation Model Orchestrationcited by 0
  26. 2026ARGS: Auto-Regressive Gaussian Splatting via Parallel Progressive Next-Scale Predictioncited by 0
  27. 2026SEM-ROVER: Semantic Voxel-Guided Diffusion for Large-Scale Driving Scene Generationcited by 0
  28. 2026VideoMDM: Towards 3D Human Motion Generation From 2D Supervisioncited by 0
  29. 20264DHumanDiff: Direct Text-to-4DGS Generation for Consistent 360-Degree Dynamic Humanscited by 0
  30. 2026Visibility-Guided Structured Measure Flow for Class-Conditioned 3D Gaussian Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Sampling 3D Gaussian Scenes in Seconds with Latent Diffusion Models
  2. 2025Generative Gaussian Splatting: Generating 3D Scenes with Video Diffusion Priors
  3. 2024LT3SD: Latent Trees for 3D Scene Diffusion
  4. 2026Generative 3D Gaussians with Learned Density Control
  5. 2024Atlas Gaussians Diffusion for 3D Generation
  6. 2024Baking Gaussian Splatting into Diffusion Denoiser for Fast and Scalable Single-stage Image-to-3D Generation and Reconstruction
  7. 2024LN3DIFF++: Scalable Latent Neural Fields Diffusion for Speedy 3D Generation
  8. 2024GaussianCube: A Structured and Explicit Radiance Representation for 3D Generative Modeling