mlsanityData to 7 Oct 2026

Go with the Flows: Mixtures of Normalizing Flows for Point Cloud Generation and Reconstruction

arXiv 2106.03135 ↗First posted 6 Jun 2021cs.CVcited by 16

Not on the map. Why?

Abstract (arXiv)

Recently normalizing flows (NFs) have demonstrated state-of-the-art performance on modeling 3D point clouds while allowing sampling with arbitrary resolution at inference time. However, these flow-based models still require long training times and large models for representing complicated geometries. This work enhances their representational power by applying mixtures of NFs to point clouds. We show that in this more general framework each component learns to specialize in a particular subregion of an object in a completely unsupervised fashion. By instantiating each mixture component with a comparatively small NF we generate point clouds with improved details compared to single-flow-based models while using fewer parameters and considerably reducing the inference runtime. We further demonstrate that by adding data augmentation, individual mixture components can learn to specialize in a semantically meaningful manner. We evaluate mixtures of NFs on generation, autoencoding and single-view reconstruction based on the ShapeNet dataset.

Key facts. Cited by 16 papers in the map (data through 7 Oct 2026). Builds on 43 of its 59 references that are in the map. First posted on arXiv 6 Jun 2021.

Builds on

43 of 59 references.

References · 43

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  3. 2016Image-to-Image Translation with Conditional Adversarial Networkscited by 4,510
  4. 2018Neural Ordinary Differential Equationscited by 3,733
  5. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,880
  6. 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,286
  7. 2017Improved Training of Wasserstein GANscited by 2,223
  8. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,220
  9. 2015Variational Inference with Normalizing Flowscited by 2,161
  10. 2016Density estimation using Real NVPcited by 2,029
  11. 2014NICE: Non-linear Independent Components Estimationcited by 1,052
  12. 2016A Point Set Generation Network for 3D Object Reconstruction from a Single Imagecited by 972
31 more
  1. 2021Diffusion Probabilistic Models for 3D Point Cloud Generationcited by 734
  2. 2018Pixel2Mesh: Generating 3D Mesh Models from Single RGB Imagescited by 649
  3. 2016Improving Variational Inference with Inverse Autoregressive Flowcited by 636
  4. 20163D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstructioncited by 587
  5. 2018FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Modelscited by 584
  6. 2019Deep Learning for 3D Point Clouds: A Surveycited by 557
  7. 2018AtlasNet: A Papier-Mach'e Approach to Learning 3D Surface Generationcited by 539
  8. 2019PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flowscited by 487
  9. 2020Reliable Fidelity and Diversity Metrics for Generative Modelscited by 425
  10. 2019What Do Single-view 3D Reconstruction Networks Learn?cited by 259
  11. 2020Learning Gradient Fields for Shape Generationcited by 245
  12. 2020SRFlow: Learning the Super-Resolution Space with Normalizing Flowcited by 209
  13. 2016Variational Lossy Autoencodercited by 175
  14. 20193D Point Cloud Generative Adversarial Network Based on Tree Structured Graph Convolutionscited by 173
  15. 2019Guided Image Generation with Conditional Invertible Neural Networkscited by 147
  16. 2020Flows for simultaneous manifold learning and density estimationcited by 117
  17. 2018PointGrow: Autoregressively Learned Point Cloud Generation with Self-Attentioncited by 105
  18. 2021Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networkscited by 101
  19. 2020SoftFlow: Probabilistic Framework for Normalizing Flow on Manifoldscited by 97
  20. 2018Multiresolution Tree Networks for 3D Point Cloud Processingcited by 91
  21. 2021Variational Transformer Networks for Layout Generationcited by 88
  22. 2020Discrete Point Flow Networks for Efficient Point Cloud Generationcited by 87
  23. 2019Semi-Supervised Learning with Normalizing Flowscited by 76
  24. 2019C-Flow: Conditional Generative Flow Models for Images and 3D Point Cloudscited by 50
  25. 2021DeFlow: Learning Complex Image Degradations from Unpaired Data with Conditional Flowscited by 30
  26. 2018Multi-chart Generative Surface Modelingcited by 26
  27. 2019A RAD approach to deep mixture modelscited by 18
  28. 2020Variational Mixture of Normalizing Flowscited by 15
  29. 2020HyperFlow: Representing 3D Objects as Surfacescited by 7
  30. 2020ChartPointFlow for Topology-Aware 3D Point Cloud Generationcited by 6
  31. 2019Conditional Invertible Flow for Point Cloud Generationcited by 3

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 5; 2023: 6; 2024: 2; 2025: 2.

16 papers.

Sort:

By area:

  1. 2022SDF-StyleGAN: Implicit SDF-Based StyleGAN for 3D Shape Generationcited by 92
  2. 2023Sketch and Text Guided Diffusion Model for Colored Point Cloud Generationcited by 36
  3. 2023GECCO: Geometrically-Conditioned Point Diffusion Modelscited by 24
  4. 2022PD-Flow: A Point Cloud Denoising Framework with Normalizing Flowscited by 19
  5. 2021PU-Flow: a Point Cloud Upsampling Network with Normalizing Flowscited by 17
  6. 2023PSRFlow: Probabilistic Super Resolution with Flow-Based Models for Scientific Datacited by 8
  7. 2023Topology-Matching Normalizing Flows for Out-of-Distribution Detection in Robot Learningcited by 7
  8. 2022Flow-based GAN for 3D Point Cloud Generation from a Single Imagecited by 4
  9. 2023Point Cloud Part Editing: Segmentation, Generation, Assembly, and Selectioncited by 4
  10. 2024FFHFlow: Diverse and Uncertainty-Aware Dexterous Grasp Generation via Flow Variational Inferencecited by 3
  11. 2022SPA-VAE: Similar-Parts-Assignment for Unsupervised 3D Point Cloud Generationcited by 2
  12. 2022ManiFlow: Implicitly Representing Manifolds with Normalizing Flowscited by 2
4 more
  1. 2023Diverse Shape Completion via Style Modulated Generative Adversarial Networkscited by 2
  2. 2024Marginalization Consistent Probabilistic Forecasting of Irregular Time Series via Mixture of Separable flowscited by 0
  3. 2025Classifier Weighted Mixture modelscited by 0
  4. 2025NF-SLAM: Effective, Normalizing Flow-supported Neural Field representations for object-level visual SLAM in automotive applicationscited by 0