mlsanityData to 14 Oct 2026

A generic diffusion-based approach for 3D human pose prediction in the wild

arXiv 2210.05669 ↗First posted 11 Oct 2022cs.CV cs.HC cs.ROcited by 33

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

Abstract (arXiv)

Predicting 3D human poses in real-world scenarios, also known as human pose forecasting, is inevitably subject to noisy inputs arising from inaccurate 3D pose estimations and occlusions. To address these challenges, we propose a diffusion-based approach that can predict given noisy observations. We frame the prediction task as a denoising problem, where both observation and prediction are considered as a single sequence containing missing elements (whether in the observation or prediction horizon). All missing elements are treated as noise and denoised with our conditional diffusion model. To better handle long-term forecasting horizon, we present a temporal cascaded diffusion model. We demonstrate the benefits of our approach on four publicly available datasets (Human3.6M, HumanEva-I, AMASS, and 3DPW), outperforming the state-of-the-art. Additionally, we show that our framework is generic enough to improve any 3D pose prediction model as a pre-processing step to repair their inputs and a post-processing step to refine their outputs. The code is available online: urlhttps://github.com/vita-epfl/DePOSit.

Key facts. Cited by 33 papers in the map (data through 14 Oct 2026). Placed in 3D, point cloud and scene generation, one of 14 directions in Diffusion models. Builds on 48 of its 66 references that are in the map. First posted on arXiv 11 Oct 2022.

Builds on

48 of 66 references.

Closest ancestors · 1

Same direction, newest first

  1. 2022RePaint: Inpainting using Denoising Diffusion Probabilistic Models24 Jan

Other references · 47

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  4. 2020Denoising Diffusion Implicit Modelscited by 9,781
  5. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
  6. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,355
  7. 2021Improved Denoising Diffusion Probabilistic Modelscited by 3,761
  8. 2019AMASS: Archive of Motion Capture as Surface Shapescited by 1,236
  9. 2021CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputationcited by 582
  10. 2021Action-Conditioned 3D Human Motion Synthesis with Transformer VAEcited by 501
  11. 2017On human motion prediction using recurrent neural networkscited by 338
  12. 2019Learning Trajectory Dependencies for Human Motion Predictioncited by 246
35 more
  1. 2015Structural-RNN: Deep Learning on Spatio-Temporal Graphscited by 234
  2. 2016Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoderscited by 220
  3. 2021Stochastic Scene-Aware Motion Predictioncited by 203
  4. 2020Long-term Human Motion Prediction with Scene Contextcited by 192
  5. 2020DLow: Diversifying Latent Flows for Diverse Human Motion Predictioncited by 185
  6. 2017HP-GAN: Probabilistic 3D human motion prediction via GANcited by 176
  7. 2018Crowd-Robot Interaction: Crowd-aware Robot Navigation with Attention-based Deep Reinforcement Learningcited by 175
  8. 2020History Repeats Itself: Human Motion Prediction via Motion Attentioncited by 165
  9. 2020A Spatio-temporal Transformer for 3D Human Motion Predictioncited by 133
  10. 2020Dynamic Multiscale Graph Neural Networks for 3D Skeleton-Based Human Motion Predictioncited by 132
  11. 2020Human Trajectory Forecasting in Crowds: A Deep Learning Perspectivecited by 129
  12. 2018Convolutional Sequence to Sequence Model for Human Dynamicscited by 115
  13. 2018Human Motion Prediction via Spatio-Temporal Inpaintingcited by 114
  14. 2017The Pose Knows: Video Forecasting by Generating Pose Futurescited by 88
  15. 2018MT-VAE: Learning Motion Transformations to Generate Multimodal Human Dynamicscited by 86
  16. 2019Context-aware Human Motion Predictioncited by 84
  17. 2021Space-Time-Separable Graph Convolutional Network for Pose Forecastingcited by 84
  18. 2019Diverse Trajectory Forecasting with Determinantal Point Processescited by 80
  19. 2017DeLiGAN : Generative Adversarial Networks for Diverse and Limited Datacited by 77
  20. 2017Learning Human Motion Models for Long-term Predictionscited by 67
  21. 2021Generating Smooth Pose Sequences for Diverse Human Motion Predictioncited by 66
  22. 2022Progressively Generating Better Initial Guesses Towards Next Stages for High-Quality Human Motion Predictioncited by 66
  23. 2018Accurate and Diverse Sampling of Sequences based on a "Best of Many" Sample Objectivecited by 56
  24. 2018Action-Agnostic Human Pose Forecastingcited by 56
  25. 2020Socially and Contextually Aware Human Motion and Pose Forecastingcited by 51
  26. 2022Spatio-Temporal Gating-Adjacency GCN for Human Motion Predictioncited by 48
  27. 2021Pose Transformers (POTR): Human Motion Prediction with Non-Autoregressive Transformerscited by 46
  28. 2021Vehicle trajectory prediction works, but not everywherecited by 39
  29. 2021Are socially-aware trajectory prediction models really socially-aware?cited by 32
  30. 2022Motron: Multimodal Probabilistic Human Motion Forecastingcited by 30
  31. 2019Disentangling Human Dynamics for Pedestrian Locomotion Forecasting with Noisy Supervisioncited by 27
  32. 2017Forecasting Human Dynamics from Static Imagescited by 17
  33. 2020Pedestrian Intention Prediction: A Multi-task Perspectivecited by 16
  34. 2022Pedestrian 3D Bounding Box Predictioncited by 11
  35. 2021A Shared Representation for Photorealistic Driving Simulatorscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 1; 2023: 10; 2024: 11; 2025: 6; 2026: 5.

33 papers.

Sort:

By area:

  1. 2023MotionDiffuser: Controllable Multi-Agent Motion Prediction using Diffusioncited by 213
  2. 2023DiffTalk: Crafting Diffusion Models for Generalized Audio-Driven Portraits Animationcited by 136
  3. 2022BeLFusion: Latent Diffusion for Behavior-Driven Human Motion Predictioncited by 82
  4. 2023Multimodal Motion Conditioned Diffusion Model for Skeleton-based Video Anomaly Detectioncited by 42
  5. 2023Social-Transmotion: Promptable Human Trajectory Predictioncited by 33
  6. 2024Multi-agent Long-term 3D Human Pose Forecasting via Interaction-aware Trajectory Conditioningcited by 30
  7. 2024Diffusion Model is a Good Pose Estimator from 3D RF-Visioncited by 22
  8. 2024Multi-Transmotion: Pre-trained Model for Human Motion Predictioncited by 21
  9. 2023Modiff: Action-Conditioned 3D Motion Generation with Denoising Diffusion Probabilistic Modelscited by 17
  10. 2023TransFusion: A Practical and Effective Transformer-based Diffusion Model for 3D Human Motion Predictioncited by 17
  11. 2025Nonisotropic Gaussian Diffusion for Realistic 3D Human Motion Predictioncited by 12
  12. 2023Toward Reliable Human Pose Forecasting with Uncertaintycited by 11
21 more
  1. 2023GazeMoDiff: Gaze-guided Diffusion Model for Stochastic Human Motion Predictioncited by 10
  2. 2023DiffHPE: Robust, Coherent 3D Human Pose Lifting with Diffusioncited by 9
  3. 2024MotionMap: Representing Multimodality in Human Pose Forecastingcited by 7
  4. 2024Certified Human Trajectory Predictioncited by 6
  5. 2024Controllable Diverse Sampling for Diffusion Based Motion Behavior Forecastingcited by 3
  6. 2025mmPred: Radar-based Human Motion Prediction in the Darkcited by 3
  7. 2024Diffusion-based Graph Generative Methodscited by 2
  8. 2024PAFUSE: Part-based Diffusion for 3D Whole-Body Pose Estimationcited by 1
  9. 2025HHI-Assist: A Dataset and Benchmark of Human-Human Interaction in Physical Assistance Scenariocited by 1
  10. 2025MMCM: Multimodality-aware Metric using Clustering-based Modes for Probabilistic Human Motion Predictioncited by 1
  11. 2026Ego-Human Motion Prediction with 3D-Aware LLMcited by 1
  12. 2023Personalized Pose Forecastingcited by 0
  13. 2024Sparse Prototype Network for Explainable Pedestrian Behavior Predictioncited by 0
  14. 2024Relation Learning and Aggregate-attention for Multi-person Motion Predictioncited by 0
  15. 2024UnityGraph: Unified Learning of Spatio-temporal features for Multi-person Motion Predictioncited by 0
  16. 2025Scriboora: Rethinking Human Pose Forecastingcited by 0
  17. 2025SMamDiff: Spatial Mamba for Stochastic Human Motion Predictioncited by 0
  18. 2026Residual Flow Matching with Dynamic Cross-Interaction for 3D Multi-Person Motion Predictioncited by 0
  19. 2026Capturing Uncertainty in Human Motion for Representation Learning in Soccercited by 0
  20. 2026Multi-Person Human Motion Forecasting in Complex Scenescited by 0
  21. 2026Everybody Tracking Every Bodycited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024D3PRefiner: A Diffusion-based Denoise Method for 3D Human Pose Refinement
  2. 2022DiffPose: Multi-hypothesis Human Pose Estimation using Diffusion models
  3. 2026Flexible Geometric Guidance for Probabilistic Human Pose Estimation with Diffusion Models
  4. 2022DiffPose: Toward More Reliable 3D Pose Estimation
  5. 2026One-Shot Novel View and Pose Human Image Synthesis via 3D Prior Guided Diffusion Model
  6. 2026Unsupervised 3D Human Pose Estimation via Conditional Multi-view Ancestral Sampling
  7. 2025DPoser-X: Diffusion Model as Robust 3D Whole-body Human Pose Prior
  8. 2024Di2Pose: Discrete Diffusion Model for Occluded 3D Human Pose Estimation