# Motion Capture is Not the Target Domain: Scaling Synthetic Data for Learning Motion Representations

arXiv: https://arxiv.org/abs/2602.11064 · HTML: https://mlsanity.com/paper/2602.11064/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 29 of its 53 references that are in the map.
- First posted on arXiv 11 Feb 2026.

## Builds on (most cited first, 29 of 29)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [ImageBind: One Embedding Space To Bind Them All](https://mlsanity.com/paper/2305.05665/) (2023)
- [MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model](https://mlsanity.com/paper/2208.15001/) (2022)
- [TEMOS: Generating diverse human motions from textual descriptions](https://mlsanity.com/paper/2204.14109/) (2022)
- [Action2Motion: Conditioned Generation of 3D Human Motions](https://mlsanity.com/paper/2007.15240/) (2020)
- [Sim-to-Real Transfer in Deep Reinforcement Learning for Robotics: a Survey](https://mlsanity.com/paper/2009.13303/) (2020)
- [Motion-X: A Large-scale 3D Expressive Whole-body Human Motion Dataset](https://mlsanity.com/paper/2307.00818/) (2023)
- [EgoBody: Human Body Shape and Motion of Interacting People from Head-Mounted Devices](https://mlsanity.com/paper/2112.07642/) (2021)
- [Deep Learning for Sensor-based Human Activity Recognition: Overview, Challenges and Opportunities](https://mlsanity.com/paper/2001.07416/) (2020)
- [Machine Learning for Synthetic Data Generation: A Review](https://mlsanity.com/paper/2302.04062/) (2023)
- [Human Motion Generation: A Survey](https://mlsanity.com/paper/2307.10894/) (2023)
- [BeLFusion: Latent Diffusion for Behavior-Driven Human Motion Prediction](https://mlsanity.com/paper/2211.14304/) (2022)
- [Self-supervised Learning for Human Activity Recognition Using 700,000 Person-days of Wearable Data](https://mlsanity.com/paper/2206.02909/) (2022)
- [Single Motion Diffusion](https://mlsanity.com/paper/2302.05905/) (2023)
- [LaMP: Language-Motion Pretraining for Motion Generation, Retrieval, and Captioning](https://mlsanity.com/paper/2410.07093/) (2024)
- [How Simulation Helps Autonomous Driving:A Survey of Sim2real, Digital Twins, and Parallel Intelligence](https://mlsanity.com/paper/2305.01263/) (2023)
- [Synthetic Datasets for Autonomous Driving: A Survey](https://mlsanity.com/paper/2304.12205/) (2023)
- [Combining Recurrent Neural Networks and Adversarial Training for Human Motion Synthesis and Control](https://mlsanity.com/paper/1806.08666/) (2018)
- [Demystifying Synthetic Data in LLM Pre-training: A Systematic Study of Scaling Laws, Benefits, and Pitfalls](https://mlsanity.com/paper/2510.01631/) (2025)
- [DiffDance: Cascaded Human Motion Diffusion Model for Dance Generation](https://mlsanity.com/paper/2308.02915/) (2023)
- [RelCon: Relative Contrastive Learning for a Motion Foundation Model for Wearable Data](https://mlsanity.com/paper/2411.18822/) (2024)
- [HiT-DVAE: Human Motion Generation via Hierarchical Transformer Dynamical VAE](https://mlsanity.com/paper/2204.01565/) (2022)
- [AAMDM: Accelerated Auto-regressive Motion Diffusion Model](https://mlsanity.com/paper/2401.06146/) (2023)
- [Self-Supervised Learning of Time Series Representation via Diffusion Process and Imputation-Interpolation-Forecasting Mask](https://mlsanity.com/paper/2405.05959/) (2024)
- [Generalizable Domain Adaptation for Sim-and-Real Policy Co-Training](https://mlsanity.com/paper/2509.18631/) (2025)
- [LEAD: Latent Realignment for Human Motion Diffusion](https://mlsanity.com/paper/2410.14508/) (2024)
- [MotionDreamer: One-to-Many Motion Synthesis with Localized Generative Masked Transformer](https://mlsanity.com/paper/2504.08959/) (2025)
- [Motion Generation: A Survey of Generative Approaches and Benchmarks](https://mlsanity.com/paper/2507.05419/) (2025)
- [Exploring the Impact of Synthetic Data on Human Gesture Recognition Tasks Using GANs](https://mlsanity.com/paper/2412.06389/) (2024)

## Abstract (arXiv)

Synthetic data offers a compelling path to scalable pretraining when real-world data is scarce, but models pretrained on synthetic data often fail to transfer reliably to deployment settings. We study this problem in full-body human motion, where large-scale data collection is infeasible but essential for wearable-based Human Activity Recognition (HAR), and where synthetic motion can be generated from motion-capture-derived representations. We pretrain motion time-series models using such synthetic data and evaluate their transfer across diverse downstream HAR tasks. Our results show that synthetic pretraining improves generalisation when mixed with real data or scaled sufficiently. We also demonstrate that large-scale motion-capture pretraining yields only marginal gains due to domain mismatch with wearable signals, clarifying key sim-to-real challenges and the limits and opportunities of synthetic motion data for transferable HAR representations.