mlsanityData to 7 Oct 2026

FLAG: Flow-based 3D Avatar Generation from Sparse Observations

arXiv 2203.05789 ↗First posted 11 Mar 2022cs.CV cs.LGcited by 54

Not on the map. Why?

Abstract (arXiv)

To represent people in mixed reality applications for collaboration and communication, we need to generate realistic and faithful avatar poses. However, the signal streams that can be applied for this task from head-mounted devices (HMDs) are typically limited to head pose and hand pose estimates. While these signals are valuable, they are an incomplete representation of the human body, making it challenging to generate a faithful full-body avatar. We address this challenge by developing a flow-based generative model of the 3D human body from sparse observations, wherein we learn not only a conditional distribution of 3D human pose, but also a probabilistic mapping from observations to the latent space from which we can generate a plausible pose along with uncertainty estimates for the joints. We show that our approach is not only a strong predictive model, but can also act as an efficient pose prior in different optimization settings where a good initial latent code plays a major role.

Key facts. Cited by 54 papers in the map (data through 7 Oct 2026). Builds on 34 of its 56 references that are in the map. First posted on arXiv 11 Mar 2022.

Builds on

34 of 56 references.

References · 34

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2013Auto-Encoding Variational Bayescited by 3,980
  3. 2021On the Opportunities and Risks of Foundation Modelscited by 3,961
  4. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  5. 2015Variational Inference with Normalizing Flowscited by 2,161
  6. 2016Density estimation using Real NVPcited by 2,029
  7. 2019Expressive Body Capture: 3D Hands, Face, and Body from a Single Imagecited by 1,600
  8. 2019AMASS: Archive of Motion Capture as Surface Shapescited by 1,208
  9. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,186
  10. 2017End-to-end Recovery of Human Shape and Posecited by 883
  11. 2016Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Imagecited by 692
  12. 2019Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loopcited by 625
22 more
  1. 2015Generating Sentences from a Continuous Spacecited by 607
  2. 2020PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitizationcited by 521
  3. 2021HuMoR: 3D Human Motion Model for Robust Pose Estimationcited by 319
  4. 2020STAR: Sparse Trained Articulated Human Body Regressorcited by 194
  5. 2021MeshTalk: 3D Face Animation from Speech using Cross-Modality Disentanglementcited by 189
  6. 2021Character Controllers Using Motion VAEscited by 182
  7. 2021Probabilistic Modeling for Human Mesh Recoverycited by 152
  8. 2021Human POSEitioning System (HPS): 3D Human Pose Estimation and Self-localization in Large Scenes from Body-Mounted Sensorscited by 137
  9. 2020We are More than Our Joints: Predicting how 3D Bodies Movecited by 115
  10. 2020Weakly Supervised 3D Human Pose and Shape Reconstruction with Normalizing Flowscited by 94
  11. 2020Contact and Human Dynamics from Monocular Videocited by 94
  12. 2019Preventing Posterior Collapse with delta-VAEscited by 86
  13. 2019Monocular 3D Human Pose Estimation by Generation and Ordinal Rankingcited by 76
  14. 2020Neural Descent for Visual 3D Human Pose and Shapecited by 60
  15. 20203D Multi-bodies: Fitting Sets of Plausible 3D Human Models to Ambiguous Image Datacited by 50
  16. 2021THUNDR: Transformer-based 3D HUmaN Reconstruction with Markerscited by 50
  17. 2021LoBSTr: Real-time Lower-body Pose Prediction from Sparse Upper-body Tracking Signalscited by 45
  18. 2021Probabilistic 3D Human Shape and Pose Estimation from Multiple Unconstrained Images in the Wildcited by 40
  19. 2020Probabilistic Tracklet Scoring and Inpainting for Multiple Object Trackingcited by 36
  20. 2020HoloLens 2 Research Mode as a Tool for Computer Vision Researchcited by 34
  21. 2021SOMA: Solving Optical Marker-Based MoCap Automaticallycited by 30
  22. 2021TrajeVAE: Controllable Human Motion Generation from Trajectoriescited by 14

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 4; 2023: 15; 2024: 16; 2025: 13; 2026: 6.

54 papers.

Sort:

By area:

  1. 2022MoFusion: A Framework for Denoising-Diffusion-based Motion Synthesiscited by 218
  2. 2023Avatars Grow Legs: Generating Smooth Human Motion from Sparse Tracking Inputs with Diffusion Modelcited by 104
  3. 2022Ego-Body Pose Estimation via Ego-Head Pose Estimationcited by 93
  4. 2022QuestSim: Human Motion Tracking from Sparse Sensors with Simulated Avatarscited by 74
  5. 2024ReGenNet: Towards Human Action-Reaction Synthesiscited by 53
  6. 2023BoDiffusion: Diffusing Sparse Observations for Full-Body Human Motion Synthesiscited by 44
  7. 2023Realistic Full-Body Tracking from Sparse Observations via Joint-Level Modelingcited by 41
  8. 2023EgoPoser: Robust Real-Time Egocentric Pose Estimation from Sparse and Intermittent Observations Everywherecited by 31
  9. 2024Mocap Everyone Everywhere: Lightweight Motion Capture With Smartwatches and a Head-Mounted Cameracited by 30
  10. 2023HuManiFlow: Ancestor-Conditioned Normalising Flows on SO(3) Manifolds for Human Pose and Shape Distribution Estimationcited by 23
  11. 2024Look Ma, no markers: holistic performance capture without the hasslecited by 23
  12. 2024HMD-Poser: On-Device Real-time Human Motion Tracking from Scalable Sparse Observationscited by 22
38 more
  1. 2023SparsePoser: Real-time Full-body Motion Reconstruction from Sparse Datacited by 20
  2. 2025Challenges and Trends in Egocentric Vision: A Surveycited by 20
  3. 2024Stratified Avatar Generation from Sparse Observationscited by 19
  4. 2023HMD-NeMo: Online 3D Avatar Motion Generation From Sparse Observationscited by 17
  5. 2023Fusing Monocular Images and Sparse IMU Signals for Real-time Human Motion Capturecited by 17
  6. 2024Real-Time Simulated Avatar from Head-Mounted Sensorscited by 17
  7. 2024A Unified Diffusion Framework for Scene-aware Human Motion Estimation from Sparse Signalscited by 15
  8. 2025HERO: Human Reaction Generation from Videoscited by 15
  9. 2024SimpleEgo: Predicting Probabilistic Body Pose from Egocentric Camerascited by 12
  10. 2024DivaTrack: Diverse Bodies and Motions from Acceleration-Enhanced Three-Point Trackerscited by 11
  11. 2024EnvPoser: Environment-aware Realistic Human Motion Estimation from Sparse Observations with Uncertainty Modelingcited by 11
  12. 2023Physics-based Motion Retargeting from Sparse Inputscited by 9
  13. 2023MVControl: Adding Conditional Control to Multi-view Diffusion for Controllable Text-to-3D Generationcited by 8
  14. 2024A Survey on 3D Egocentric Human Pose Estimationcited by 8
  15. 2025ARFlow: Human Action-Reaction Flow Matching with Physical Guidancecited by 7
  16. 2023MOVIN: Real-time Motion Capture using a Single LiDARcited by 6
  17. 2023ET3D: Efficient Text-to-3D Generation via Multi-View Distillationcited by 6
  18. 2024Generalizable Face Landmarking Guided by Conditional Face Warpingcited by 6
  19. 2025EventEgo3D++: 3D Human Motion Capture from a Head-Mounted Event Cameracited by 6
  20. 2024StackFLOW: Monocular Human-Object Reconstruction by Stacked Normalizing Flow with Offsetcited by 5
  21. 2023Improving the Robustness of 3D Human Pose Estimation: A Benchmark and Learning from Noisy Inputcited by 4
  22. 2025Progressive Inertial Poser: Progressive Real-Time Kinematic Chain Estimation for 3D Full-Body Pose from Three IMU Sensorscited by 4
  23. 2024DragPoser: Motion Reconstruction from Variable Sparse Tracking Signals via Latent Space Optimizationcited by 3
  24. 2024XR-MBT: Multi-modal Full Body Tracking for XR through Self-Supervision with Learned Depth Point Cloud Registrationcited by 3
  25. 2022MVP-Human Dataset for 3D Human Avatar Reconstruction from Unconstrained Framescited by 2
  26. 2023Utilizing Task-Generic Motion Prior to Recover Full-Body Motion from Very Sparse Signalscited by 2
  27. 2025DiffCap: Diffusion-based Real-time Human Motion Capture using Sparse IMUs and a Monocular Cameracited by 2
  28. 2026Ground Reaction Inertial Poser: Physics-based Human Motion Capture from Sparse IMUs and Insole Pressure Sensorscited by 2
  29. 2024Biometrics in Extended Reality: A Reviewcited by 1
  30. 2025SSD-Poser: Avatar Pose Estimation with State Space Duality from Sparse Observationscited by 1
  31. 2023QPoser: Quantized Explicit Pose Prior Modeling for Controllable Pose Generationcited by 0
  32. 2025Efficient 3D Full-Body Motion Generation from Sparse Tracking Inputs with Temporal Windowscited by 0
  33. 2025MAGE:A Multi-stage Avatar Generator with Sparse Observationscited by 0
  34. 2025Zero-shot Human Pose Estimation using Diffusion-based Inverse solverscited by 0
  35. 2025Shape-aware Inertial Poser: Motion Tracking for Humans with Diverse Shapes Using Sparse Inertial Sensorscited by 0
  36. 2025Mem-MLP: Real-Time 3D Human Motion Generation from Sparse Inputscited by 0
  37. 2025OmniEgoCap: Camera-Agnostic Sequence-Level Egocentric Motion Reconstructioncited by 0
  38. 2026Int3DNet: Scene-Motion Cross Attention Network for 3D Intention Prediction in Mixed Realitycited by 0

50 most cited of 54 All 54 (JSON)