mlsanityData to 7 Oct 2026

DiffPose: Multi-hypothesis Human Pose Estimation using Diffusion models

arXiv 2211.16487 ↗First posted 29 Nov 2022cs.CVcited by 65

On the map: Diffusion models → 3D, point cloud and scene generation

Abstract (arXiv)

Traditionally, monocular 3D human pose estimation employs a machine learning model to predict the most likely 3D pose for a given input image. However, a single image can be highly ambiguous and induces multiple plausible solutions for the 2D-3D lifting step which results in overly confident 3D pose predictors. To this end, we propose DiffPose, a conditional diffusion model, that predicts multiple hypotheses for a given input image. In comparison to similar approaches, our diffusion model is straightforward and avoids intensive hyperparameter tuning, complex network structures, mode collapse, and unstable training. Moreover, we tackle a problem of the common two-step approach that first estimates a distribution of 2D joint locations via joint-wise heatmaps and consecutively approximates them based on first- or second-moment statistics. Since such a simplification of the heatmaps removes valid information about possibly correct, though labeled unlikely, joint locations, we propose to represent the heatmaps as a set of 2D joint candidate samples. To extract information about the original distribution from these samples we introduce our embedding transformer that conditions the diffusion model. Experimentally, we show that DiffPose slightly improves upon the state of the art for multi-hypothesis pose estimation for simple poses and outperforms it by a large margin for highly ambiguous poses.

Key facts. Cited by 65 papers in the map (data through 7 Oct 2026). Placed in 3D, point cloud and scene generation, one of 14 directions in Diffusion models. Builds on 37 of its 56 references that are in the map. First posted on arXiv 29 Nov 2022.

Builds on

37 of 56 references.

References · 37

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  4. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,539
  5. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,322
  6. 2022Generative Adversarial Networkscited by 5,603
  7. 2021Improved Denoising Diffusion Probabilistic Modelscited by 3,736
  8. 2019Deep High-Resolution Representation Learning for Human Pose Estimationcited by 1,339
  9. 2017End-to-end Recovery of Human Shape and Posecited by 883
  10. 2016Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Imagecited by 692
  11. 2019Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loopcited by 625
  12. 2019VIBE: Video Inference for Human Body Pose and Shape Estimationcited by 615
25 more
  1. 2021CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputationcited by 577
  2. 2016Monocular 3D Human Pose Estimation In The Wild Using Improved CNN Supervisioncited by 493
  3. 2017A simple yet effective baseline for 3d human pose estimationcited by 401
  4. 2020HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimationcited by 304
  5. 2019Semantic Graph Convolutional Networks for 3D Human Pose Regressioncited by 204
  6. 2021Probabilistic Modeling for Human Mesh Recoverycited by 152
  7. 2021MHFormer: Multi-Hypothesis Transformer for 3D Human Pose Estimationcited by 151
  8. 20163D Human Pose Estimation = 2D Pose Estimation + Matchingcited by 112
  9. 2021D2C: Diffusion-Denoising Models for Few-shot Conditional Generationcited by 96
  10. 2020Weakly Supervised 3D Human Pose and Shape Reconstruction with Normalizing Flowscited by 94
  11. 2019Generating Multiple Hypotheses for 3D Human Pose Estimation with Mixture Density Networkcited by 92
  12. 2019RepNet: Weakly Supervised Training of an Adversarial Reprojection Network for 3D Human Pose Estimationcited by 85
  13. 2021Probabilistic Monocular 3D Human Pose Estimation with Normalizing Flowscited by 84
  14. 2020Cascaded deep monocular 3D human pose estimation with evolutionary training datacited by 80
  15. 2019Monocular 3D Human Pose Estimation by Generation and Ordinal Rankingcited by 76
  16. 20163D Human Pose Estimation from a Single Image via Distance Matrix Regressioncited by 73
  17. 2020CanonPose: Self-Supervised Monocular 3D Human Pose Estimation in the Wildcited by 65
  18. 20203D Multi-bodies: Fitting Sets of Plausible 3D Human Models to Ambiguous Image Datacited by 50
  19. 2019In the Wild Human Pose Estimation Using Explicit 2D Features and Intermediate 3D Representationscited by 48
  20. 2021Adversarial Parametric Pose Priorcited by 38
  21. 2021ElePose: Unsupervised 3D Human Pose Estimation by Predicting Camera Elevation and Learning Normalizing Flows on 2D Posescited by 35
  22. 2017Generating Multiple Diverse Hypotheses for Human 3D Pose Consistent with 2D Joint Detectionscited by 30
  23. 2020GraphMDN: Leveraging graph structure and deep learning to solve inverse problemscited by 26
  24. 2020Weakly Supervised Generative Network for Multiple 3D Human Pose Hypothesescited by 25
  25. 2019Not All Parts Are Created Equal: 3D Pose Estimation by Modelling Bi-directional Dependencies of Body Partscited by 19

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 14; 2024: 23; 2025: 19; 2026: 9.

65 papers.

Sort:

By area:

  1. 2023Diffusion-Based 3D Human Pose Estimation with Multi-Hypothesis Aggregationcited by 81
  2. 2023Egocentric Whole-Body Motion Capture with FisheyeViT and Diffusion-Based Motion Refinementcited by 31
  3. 2023DiffusionPoser: Real-time Human Motion Reconstruction From Arbitrary Sparse Sensors Using Autoregressive Diffusioncited by 29
  4. 2024Multi-agent Long-term 3D Human Pose Forecasting via Interaction-aware Trajectory Conditioningcited by 29
  5. 2023Parallel Vertex Diffusion for Unified Visual Groundingcited by 28
  6. 2024Motion2VecSets: 4D Latent Vector Set Diffusion for Non-rigid Shape Reconstruction and Trackingcited by 27
  7. 2024HMD^2: Environment-aware Motion Generation from Single Egocentric Head-Mounted Devicecited by 26
  8. 2023Diffusion Model for Dense Matchingcited by 21
  9. 2024Closely Interactive Human Reconstruction with Proxemics and Physics-Guided Adaptioncited by 18
  10. 2023Confronting Ambiguity in 6D Object Pose Estimation via Score-Based Diffusion on SE(3)cited by 17
  11. 2024HHMR: Holistic Hand Mesh Recovery by Enhancing the Multimodal Controllability of Graph Diffusion Modelscited by 14
  12. 2025REWIND: Real-Time Egocentric Whole-Body Motion Diffusion with Exemplar-Based Identity Conditioningcited by 14
38 more
  1. 2025DPoser-X: Diffusion Model as Robust 3D Whole-body Human Pose Priorcited by 13
  2. 2023MonoDiffusion: Self-Supervised Monocular Depth Estimation Using Diffusion Modelcited by 12
  3. 2024HandDiff: 3D Hand Pose Estimation with Diffusion on Image-Point Cloudcited by 12
  4. 2024Open-Vocabulary 3D Semantic Segmentation with Text-to-Image Diffusion Modelscited by 12
  5. 2024AddBiomechanics Dataset: Capturing the Physics of Human Motion at Scalecited by 11
  6. 2024FurniScene: A Large-scale 3D Room Dataset with Intricate Furnishing Scenescited by 10
  7. 2023Guided Diffusion from Self-Supervised Diffusion Featurescited by 9
  8. 2024Bridging the Skeleton-Text Modality Gap: Diffusion-Powered Modality Alignment for Zero-shot Skeleton-based Action Recognitioncited by 9
  9. 2025HiPART: Hierarchical Pose AutoRegressive Transformer for Occluded 3D Human Pose Estimationcited by 8
  10. 2023DiffHand: End-to-End Hand Mesh Reconstruction via Diffusion Modelscited by 7
  11. 2023Human Pose-based Estimation, Tracking and Action Recognition with Deep Learning: A Surveycited by 6
  12. 2023ManiPose: Manifold-Constrained Multi-Hypothesis 3D Human Pose Estimationcited by 6
  13. 2024DiffSF: Diffusion Models for Scene Flow Estimationcited by 6
  14. 2024UPose3D: Uncertainty-Aware 3D Human Pose Estimation with Cross-View and Temporal Cuescited by 6
  15. 2023Learning Structure-Guided Diffusion Model for 2D Human Pose Estimationcited by 5
  16. 2024Multi-hypotheses Conditioned Point Cloud Diffusion for 3D Human Reconstruction from Occluded Imagescited by 5
  17. 2024CHAMP: Conformalized 3D Human Multi-Hypothesis Pose Estimatorscited by 4
  18. 2024Utilizing Uncertainty in 2D Pose Detectors for Probabilistic 3D Human Mesh Recoverycited by 4
  19. 2024Robot Shape and Location Retention in Video Generation Using Diffusion Modelscited by 3
  20. 2024ViMo: Generating Motions from Casual Videoscited by 3
  21. 2026FMPose3D: monocular 3D pose estimation via flow matchingcited by 3
  22. 2023Denoising Diffusion for 3D Hand Pose Estimation from Imagescited by 2
  23. 2023Physics-Guided Human Motion Capture with Pose Probability Modelingcited by 2
  24. 2024Platypose: Calibrated Zero-Shot Multi-Hypothesis 3D Human Motion Estimationcited by 2
  25. 2024Di2Pose: Discrete Diffusion Model for Occluded 3D Human Pose Estimationcited by 2
  26. 2024CatFree3D: Category-agnostic 3D Object Detection with Diffusioncited by 2
  27. 2024CUPS: Improving Human Pose-Shape Estimators with Conformalized Deep Uncertaintycited by 2
  28. 2025NanoHTNet: Nano Human Topology Network for Efficient 3D Human Pose Estimationcited by 2
  29. 2025DiffCap: Diffusion-based Real-time Human Motion Capture using Sparse IMUs and a Monocular Cameracited by 2
  30. 2025Blind-Spot Guided Diffusion for Self-supervised Real-World Denoisingcited by 2
  31. 2025DreamPose3D: Hallucinative Diffusion with Prompt Learning for 3D Human Pose Estimationcited by 2
  32. 2025Exploiting Spatiotemporal Properties for Efficient Event-Driven Human Pose Estimationcited by 2
  33. 2026Flow Matching for Probabilistic Monocular 3D Human Pose Estimationcited by 2
  34. 2026DICArt: Advancing Category-level Articulated Object Pose Estimation in Discrete State-Spacescited by 2
  35. 2026MuPPet: Multi-person 2D-to-3D Pose Liftingcited by 2
  36. 2024Graph and Skipped Transformer: Exploiting Spatial and Temporal Modeling Capacities for Efficient 3D Human Pose Estimationcited by 1
  37. 2024PAFUSE: Part-based Diffusion for 3D Whole-Body Pose Estimationcited by 1
  38. 2025Physics-informed Ground Reaction Dynamics from Human Motion Capturecited by 1

50 most cited of 65 All 65 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2022DiffPose: Toward More Reliable 3D Pose Estimation
  2. 2023DiffHPE: Robust, Coherent 3D Human Pose Lifting with Diffusion
  3. 2023DiffPose: SpatioTemporal Diffusion Model for Video-Based Human Pose Estimation
  4. 2022A generic diffusion-based approach for 3D human pose prediction in the wild
  5. 2022DiffuPose: Monocular 3D Human Pose Estimation via Denoising Diffusion Probabilistic Model
  6. 2024D3PRefiner: A Diffusion-based Denoise Method for 3D Human Pose Refinement
  7. 2024Disentangled Diffusion-Based 3D Human Pose Estimation with Hierarchical Spatial and Temporal Denoiser
  8. 2025SDPose: Exploiting Diffusion Priors for Out-of-Domain and Robust Pose Estimation