# FastHMR: Accelerating Human Mesh Recovery via Token and Layer Merging with Diffusion Decoding

arXiv: https://arxiv.org/abs/2510.10868 · HTML: https://mlsanity.com/paper/2510.10868/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 46 of its 56 references that are in the map.
- First posted on arXiv 13 Oct 2025.

## Builds on (most cited first, 30 of 46)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Expressive Body Capture: 3D Hands, Face, and Body from a Single Image](https://mlsanity.com/paper/1904.05866/) (2019)
- [AMASS: Archive of Motion Capture as Surface Shapes](https://mlsanity.com/paper/1904.03278/) (2019)
- [Token Merging: Your ViT But Faster](https://mlsanity.com/paper/2210.09461/) (2022)
- [End-to-end Recovery of Human Shape and Pose](https://mlsanity.com/paper/1712.06584/) (2017)
- [Vision Transformers Need Registers](https://mlsanity.com/paper/2309.16588/) (2023)
- [Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image](https://mlsanity.com/paper/1607.08128/) (2016)
- [Do Vision Transformers See Like Convolutional Neural Networks?](https://mlsanity.com/paper/2108.08810/) (2021)
- [Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loop](https://mlsanity.com/paper/1909.12828/) (2019)
- [VIBE: Video Inference for Human Body Pose and Shape Estimation](https://mlsanity.com/paper/1912.05656/) (2019)
- [Executing your Commands via Motion Diffusion in Latent Space](https://mlsanity.com/paper/2212.04048/) (2022)
- [All are Worth Words: A ViT Backbone for Diffusion Models](https://mlsanity.com/paper/2209.12152/) (2022)
- [Action-Conditioned 3D Human Motion Synthesis with Transformer VAE](https://mlsanity.com/paper/2104.05670/) (2021)
- [Monocular 3D Human Pose Estimation In The Wild Using Improved CNN Supervision](https://mlsanity.com/paper/1611.09813/) (2016)
- [ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation](https://mlsanity.com/paper/2204.12484/) (2022)
- [Humans in 4D: Reconstructing and Tracking Humans with Transformers](https://mlsanity.com/paper/2305.20091/) (2023)
- [Learning 3D Human Dynamics from Video](https://mlsanity.com/paper/1812.01601/) (2018)
- [Common Diffusion Noise Schedules and Sample Steps are Flawed](https://mlsanity.com/paper/2305.08891/) (2023)
- [HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimation](https://mlsanity.com/paper/2011.14672/) (2020)
- [CLIFF: Carrying Location Information in Full Frames into Human Pose and Shape Estimation](https://mlsanity.com/paper/2208.00571/) (2022)
- [BEDLAM: A Synthetic Dataset of Bodies Exhibiting Detailed Lifelike Animated Motion](https://mlsanity.com/paper/2306.16940/) (2023)
- [WHAM: Reconstructing World-grounded Humans with Accurate 3D Motion](https://mlsanity.com/paper/2312.07531/) (2023)
- [MotionBERT: A Unified Perspective on Learning Human Motion Representations](https://mlsanity.com/paper/2210.06551/) (2022)
- [Deep Patch Visual Odometry](https://mlsanity.com/paper/2208.04726/) (2022)
- [MotionLCM: Real-time Controllable Motion Generation via Latent Consistency Model](https://mlsanity.com/paper/2404.19759/) (2024)
- [Beyond Static Features for Temporally Consistent 3D Human Pose and Shape from a Video](https://mlsanity.com/paper/2011.08627/) (2020)
- [MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose Estimation in Video](https://mlsanity.com/paper/2203.00859/) (2022)
- [EMDB: The Electromagnetic Database of Global 3D Human Pose and Shape in the Wild](https://mlsanity.com/paper/2308.16894/) (2023)

## Built on by (most cited first, 4 of 4)

- [Fast SAM 3D Body: Accelerating SAM 3D Body for Real-Time Full-Body Human Mesh Recovery](https://mlsanity.com/paper/2603.15603/) (2026)
- [OA-NBV: Occlusion-Aware Next-Best-View Planning for Human-Centered Active Perception on Mobile Robots](https://mlsanity.com/paper/2603.11072/) (2026)
- [AIGaitor: Privacy-preserving and cloud-free motion analysis for everyone, using edge computing](https://mlsanity.com/paper/2605.21421/) (2026)
- [SynthGait-19K: A Physically Grounded Synthetic Video Dataset for Gait Parameter Estimation](https://mlsanity.com/paper/2609.08108/) (2026)

## Abstract (arXiv)

Recent transformer-based models for 3D Human Mesh Recovery (HMR) have achieved strong performance but often suffer from high computational cost and complexity due to deep transformer architectures and redundant tokens. In this paper, we introduce two HMR-specific merging strategies: Error-Constrained Layer Merging (ECLM) and Mask-guided Token Merging (Mask-ToMe). ECLM selectively merges transformer layers that have minimal impact on the Mean Per Joint Position Error (MPJPE), while Mask-ToMe focuses on merging background tokens that contribute little to the final prediction. To further address the potential performance drop caused by merging, we propose a diffusion-based decoder that incorporates temporal context and leverages pose priors learned from large-scale motion capture datasets. Experiments across multiple benchmarks demonstrate that our method achieves up to 2.3x speed-up while slightly improving performance over the baseline.