3D Human Body Models, Mesh Recovery and Motion Generation
Parametric 3D models of the body, hands and face, fitted to images or generated from text.
Tier Grades
Tier A: Frontier AI labs
feasible
Body models from thousands of scans come from labs such as MPI-IS.
Tier B: Indonesian AI labs
stretch
Collecting scan data is a stretch for an Indonesian lab; fitting released models is feasible.
Tier C: Institution units such as ITB
feasible
Mesh recovery research fits a university unit, as the 7-author SMPL-X paper shows.
Tier D: Organisations such as SEMAI (PoT, PoC, PoV)
stretch
Client value is mainly in ergonomics and animation, which needs specialist work.
Where this comes from
Tiers: a) frontier AI labs, b) Indonesian AI labs, c) institution units such as ITB, d) organisations such as SEMAI delivering proof of technology, concept and value. Each rationale cites the anchors below.
Anchors
- ComputeTraining compute is not in the sources read.
- DataSMPL-X is trained from thousands of 3D scans.
- Team size (proxy)Proxy only (author count of an exemplar paper, not a headcount): 7 authors on the SMPL-X paper (arXiv 1904.05866).
- Time horizonYears.
- Shelf-life signalsText-to-motion models drive avatars and robots, Body models are learned end to end from video
- Pipeline roleTier A and research institutes build body models; Tier C applies them to motion analysis; Tier D uses them only in specialist pilots.
Where this comes from
Team size is a proxy (author counts of exemplar papers), not a headcount.
Who and What
Exemplars
Observed actors
Linked models
HY-Motion 1.03DexemplarSMPL-X and SMPLify-X3Dexemplar
HMR (Human Mesh Recovery)visionexemplarMANO3Dexemplar
CLIFFvisionproduced by
PyMAF-Xvisionproduced byPAREvisionproduced by
HybrIKvisionproduced by
InterNet (InterHand2.6M)visionproduced byExPosevisionproduced bySTAR3Dproduced byVIBEvisionproduced by
SPINvisionproduced by
SMPLifyvisionproduced byLinks
R&D Topic3D Human Body Models, Mesh Recovery and Motion Generation
ModelsHY-Motion 1.0Tencent, 3DSMPL-X and SMPLify-XMax Planck Institute for Intelligent Systems, 3DMANOMax Planck Institute for Intelligent Systems, 3DHMR (Human Mesh Recovery)UC Berkeley, vision+10 more
HY-Motion 1.0Tencent, 3DSMPL-X and SMPLify-XMax Planck Institute for Intelligent Systems, 3DMANOMax Planck Institute for Intelligent Systems, 3DHMR (Human Mesh Recovery)UC Berkeley, vision+10 moreUse Casesnone recorded
Who Runs Itnone recorded
Models 14
InterNet (InterHand2.6M)Seoul National University, Meta, visionproduced byPAREMax Planck Institute for Intelligent Systems, ETH Zurich, visionproduced bySMPLifyMicrosoft Research, Max Planck Institute for Intelligent Systems, visionproduced byCLIFFHuawei, visionproduced byExPoseMax Planck Institute for Intelligent Systems, visionproduced byHMR (Human Mesh Recovery)UC Berkeley, visionexemplarHY-Motion 1.0Tencent, 3DexemplarHybrIKShanghai Jiao Tong University, visionproduced byMANOMax Planck Institute for Intelligent Systems, 3DexemplarPyMAF-XTsinghua University, visionproduced bySMPL-X and SMPLify-XMax Planck Institute for Intelligent Systems, 3DexemplarSPINUniversity of Pennsylvania, visionproduced byPapers
- Expressive Body Capture: 3D Hands, Face, and Body from a Single ImageGeorgios Pavlakos et al., CVPR 2019 (per arXiv comment), 2019 founding
- End-to-end Recovery of Human Shape and PoseAngjoo Kanazawa et al., CVPR 2018 (per arXiv comment), 2017 founding
- Humans in 4D: Reconstructing and Tracking Humans with TransformersShubham Goel et al., ICCV 2023 (per arXiv comment), 2023 state of the art
- SMPLer-X: Scaling Up Expressive Human Pose and Shape EstimationZhongang Cai et al., arXiv preprint, 2023 state of the art
- WHAM: Reconstructing World-grounded Humans with Accurate 3D MotionSoyong Shin et al., arXiv preprint, 2023 state of the art
Where this comes from
The papers to read first: where the direction began, its current best, and a survey when one exists.
Sources
- arxiv.org/abs/1904.05866 https://arxiv.org/abs/1904.05866
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "3D Human Body Models, Mesh Recovery and Motion Generation", https://aurion.semai.id/topics/3d-human-body-modelling/, records of October 2026.