mlsanityData to 7 Oct 2026

TRACE: 5D Temporal Regression of Avatars with Dynamic Cameras in 3D Environments

arXiv 2306.02850 ↗First posted 5 Jun 2023cs.CVcited by 89

Not on the map. Why?

Abstract (arXiv)

Although the estimation of 3D human pose and shape (HPS) is rapidly progressing, current methods still cannot reliably estimate moving humans in global coordinates, which is critical for many applications. This is particularly challenging when the camera is also moving, entangling human and camera motion. To address these issues, we adopt a novel 5D representation (space, time, and identity) that enables end-to-end reasoning about people in scenes. Our method, called TRACE, introduces several novel architectural components. Most importantly, it uses two new "maps" to reason about the 3D trajectory of people over time in camera, and world, coordinates. An additional memory unit enables persistent tracking of people even during long occlusions. TRACE is the first one-stage method to jointly recover and track 3D humans in global coordinates from dynamic cameras. By training it end-to-end, and using full image information, TRACE achieves state-of-the-art performance on tracking and HPS benchmarks. The code and dataset are released for research purposes.

Key facts. Cited by 89 papers in the map (data through 7 Oct 2026). Builds on 51 of its 73 references that are in the map. First posted on arXiv 5 Jun 2023.

Builds on

51 of 73 references.

References · 51

  1. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,385
  2. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,186
  3. 2021YOLOX: Exceeding YOLO Series in 2021cited by 1,064
  4. 2017End-to-end Recovery of Human Shape and Posecited by 883
  5. 2021ByteTrack: Multi-Object Tracking by Associating Every Detection Boxcited by 812
  6. 2016Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Imagecited by 692
  7. 2016Performance Measures and a Data Set for Multi-Target, Multi-Camera Trackingcited by 685
  8. 2018Deformable ConvNets v2: More Deformable, Better Resultscited by 666
  9. 2019Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loopcited by 625
  10. 2019VIBE: Video Inference for Human Body Pose and Shape Estimationcited by 615
  11. 2020HOTA: A Higher Order Metric for Evaluating Multi-Object Trackingcited by 546
  12. 2021TrackFormer: Multi-Object Tracking with Transformerscited by 515
38 more
  1. 2016Monocular 3D Human Pose Estimation In The Wild Using Improved CNN Supervisioncited by 493
  2. 2022ViTPose: Simple Vision Transformer Baselines for Human Pose Estimationcited by 450
  3. 2020End-to-End Human Pose and Mesh Reconstruction with Transformerscited by 417
  4. 2019Tracking without bells and whistlescited by 382
  5. 2020MOT20: A benchmark for multi object tracking in crowded scenescited by 355
  6. 2020HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimationcited by 304
  7. 2021ICON: Implicit Clothed humans Obtained from Normalscited by 283
  8. 2021PyMAF: 3D Human Pose and Shape Regression with Pyramidal Mesh Alignment Feedback Loopcited by 250
  9. 2019HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimationcited by 247
  10. 2022CLIFF: Carrying Location Information in Full Frames into Human Pose and Shape Estimationcited by 237
  11. 2020Monocular, One-stage, Regression of Multiple 3D Peoplecited by 216
  12. 2022ECON: Explicit Clothed humans Optimized via Normal integrationcited by 199
  13. 2017Single-Shot Multi-Person 3D Pose Estimation From Monocular RGBcited by 195
  14. 2022Deep Patch Visual Odometrycited by 186
  15. 2022Generating Holistic 3D Human Motion from Speechcited by 180
  16. 2017CityPersons: A Diverse Dataset for Pedestrian Detectioncited by 171
  17. 2021Probabilistic Modeling for Human Mesh Recoverycited by 152
  18. 2021GLAMR: Global Occlusion-Aware Human Mesh Recovery with Dynamic Camerascited by 151
  19. 2020TAO: A Large-Scale Benchmark for Tracking Any Objectcited by 146
  20. 2021Putting People in their Place: Monocular Regression of 3D People in Depthcited by 141
  21. 2023Decoupling Human and Camera Motion from Videos in the Wildcited by 140
  22. 2020Coherent Reconstruction of Multiple Humans from a Single Imagecited by 136
  23. 2021SPEC: Seeing People in the Wild with an Estimated Cameracited by 134
  24. 20233D Human Pose Estimation via Intuitive Physicscited by 100
  25. 2019The 2019 DAVIS Challenge on VOS: Unsupervised Multi-Object Segmentationcited by 95
  26. 2019Human Mesh Recovery from Monocular Images via a Skeleton-disentangled Representationcited by 87
  27. 2018Pose Flow: Efficient Online Pose Trackingcited by 80
  28. 2021MOTSynth: How Can Synthetic Data Help Pedestrian Detection and Tracking?cited by 68
  29. 2021DeepMultiCap: Performance Capture of Multiple Characters Using Sparse Multiview Camerascited by 67
  30. 2022MIME: Human-Aware 3D Scene Generationcited by 57
  31. 2019TexturePose: Supervising Human Mesh Estimation with Texture Consistencycited by 53
  32. 20204D Human Body Capture from Egocentric Video via 3D Scene Groundingcited by 43
  33. 2019Learning Temporal Pose Estimation from Sparsely-Labeled Videoscited by 39
  34. 2022D&D: Learning Human Dynamics from Dynamic Cameracited by 39
  35. 2021JRDB-Act: A Large-scale Dataset for Spatio-temporal Action, Social Group and Activity Detectioncited by 38
  36. 2023High-Fidelity Clothed Avatar Reconstruction from a Single Imagecited by 38
  37. 2022The One Where They Reconstructed 3D Humans and Environments in TV Showscited by 30
  38. 2021Tracking People with 3D Representationscited by 22

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 7; 2024: 30; 2025: 26; 2026: 26.

89 papers.

Sort:

By area:

  1. 2023BEDLAM: A Synthetic Dataset of Bodies Exhibiting Detailed Lifelike Animated Motioncited by 229
  2. 2023WHAM: Reconstructing World-grounded Humans with Accurate 3D Motioncited by 200
  3. 2024World-Grounded Human Motion Recovery via Gravity-View Coordinatescited by 127
  4. 2023GART: Gaussian Articulated Template Modelscited by 121
  5. 2024TRAM: Global Trajectory and Motion of 3D Humans from in-the-wild Videoscited by 105
  6. 2025PromptHMR: Promptable Human Mesh Recoverycited by 70
  7. 2024RoHM: Robust Human Motion Reconstruction via Diffusioncited by 46
  8. 2025GENMO: A GENeralist Model for Human MOtioncited by 43
  9. 2024AiOS: All-in-One-Stage Expressive Human Pose and Shape Estimationcited by 41
  10. 2025Human3R: Everyone Everywhere All at Oncecited by 35
  11. 2023Human101: Training 100+FPS Human Gaussians in 100s from 1 Viewcited by 32
  12. 2024Reconstructing People, Places, and Camerascited by 31
38 more
  1. 2024WHAC: World-grounded Humans and Camerascited by 29
  2. 2025Joint Optimization for 4D Human-Scene Reconstruction in the Wildcited by 29
  3. 2024Spatial Cognition from Egocentric Video: Out of Sight, Not Out of Mindcited by 28
  4. 2024PhysPT: Physics-aware Pretrained Transformer for Estimating Human Dynamics from Monocular Videoscited by 27
  5. 2024Towards Open Domain Text-Driven Synthesis of Multi-Person Motionscited by 27
  6. 2025Uni4D: Unifying Visual Foundation Models for 4D Modeling from a Single Videocited by 27
  7. 2025BEDLAM2.0: Synthetic Humans and Cameras in Motioncited by 21
  8. 2024MultiPhys: Multi-Person Physics-aware 3D Motion Estimationcited by 18
  9. 2024Synergistic Global-space Camera and Human Reconstruction from Videoscited by 17
  10. 2024MMVP: A Multimodal MoCap Dataset with Vision and Pressure Sensorscited by 16
  11. 2024RELI11D: A Comprehensive Multimodal Human Motion Dataset and Methodcited by 16
  12. 2024COIN: Control-Inpainting Diffusion Prior for Human and Camera Motion Estimationcited by 15
  13. 2024MotionBank: A Large-scale Video Motion Benchmark with Disentangled Rule-based Annotationscited by 14
  14. 2025WorldPose: A World Cup Dataset for Global 3D Human Pose Estimationcited by 13
  15. 2025CoMotion: Concurrent Multi-person 3D Motioncited by 13
  16. 2025Reconstructing 4D Spatial Intelligence: A Surveycited by 13
  17. 2024AWOL: Analysis WithOut synthesis using Languagecited by 12
  18. 2024MultiPly: Reconstruction of Multiple People from Monocular Video in the Wildcited by 12
  19. 2024Lifting Motion to the 3D World via 2D Diffusioncited by 11
  20. 2023ProxyCap: Real-time Monocular Full-body Capture in World Space via Human-Centric Proxy-to-Motion Learningcited by 9
  21. 2025ODHSR: Online Dense 3D Reconstruction of Humans and Scenes from Monocular Videoscited by 8
  22. 2025Pulp Motion: Framing-aware multimodal camera and human motion generationcited by 8
  23. 2025Perceiving and Acting in First-Person: A Dataset and Benchmark for Egocentric Human-Object-Human Interactionscited by 7
  24. 2026UniSH: Unifying Scene and Human Reconstruction in a Feed-Forward Passcited by 7
  25. 2024Optimal-state Dynamics Estimation for Physics-based Human Motion Capture from Videoscited by 6
  26. 2024Humans as a Calibration Pattern: Dynamic 3D Scene Reconstruction from Unsynchronized and Uncalibrated Videoscited by 6
  27. 2025Geometric Neural Distance Fields for Learning Human Motion Priorscited by 6
  28. 2026DuoMo: Dual Motion Diffusion for World-Space Human Reconstructioncited by 6
  29. 2025PhysHMR: Learning Humanoid Control Policies from Vision for Physically Plausible Human Motion Reconstructioncited by 5
  30. 2023Exploring Event-based Human Pose Estimation with 3D Event Representationscited by 4
  31. 2024A Plug-and-Play Physical Motion Restoration Approach for In-the-Wild High-Difficulty Motionscited by 4
  32. 2026OnlineHMR: Video-based Online World-Grounded Human Mesh Recoverycited by 4
  33. 2026THFM: A Unified Video Foundation Model for 4D Human Perception and Beyondcited by 4
  34. 2024WildAvatar: Learning In-the-wild 3D Avatars from the Webcited by 3
  35. 2025ClimbingCap: Multi-Modal Dataset and Method for Rock Climbing in World Coordinatecited by 3
  36. 2025MetricHMSR:Metric Human Mesh and Scene Recovery from Monocular Imagescited by 3
  37. 2025DyCrowd: Towards Dynamic Crowd Reconstruction from a Large-scene Videocited by 3
  38. 2025Towards Metric-Aware Multi-Person Mesh Recovery by Jointly Optimizing Human Crowd in Camera Spacecited by 3

50 most cited of 89 All 89 (JSON)