Saved in:
Bibliographic Details
Main Authors: Fan, Junqiao, Liu, Pengfei, Rao, Haocong
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.00355
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909933866319872
author Fan, Junqiao
Liu, Pengfei
Rao, Haocong
author_facet Fan, Junqiao
Liu, Pengfei
Rao, Haocong
contents With intelligent room-side sensing and service robots widely deployed, human motion prediction (HMP) is essential for safe, proactive assistance. However, many existing HMP methods either produce a single, deterministic forecast that ignores uncertainty or rely on probabilistic models that sacrifice kinematic plausibility. Diffusion models improve the accuracy-diversity trade-off but often depend on multi-stage pipelines that are costly for edge deployment. This work focuses on how to ensure spatial-temporal coherence within a single-stage diffusion model for HMP. We introduce SMamDiff, a Spatial Mamba-based Diffusion model with two novel designs: (i) a residual-DCT motion encoding that subtracts the last observed pose before a temporal DCT, reducing the first DC component ($f=0$) dominance and highlighting informative higher-frequency cues so the model learns how joints move rather than where they are; and (ii) a stickman-drawing spatial-mamba module that processes joints in an ordered, joint-by-joint manner, making later joints condition on earlier ones to induce long-range, cross-joint dependencies. On Human3.6M and HumanEva, these coherence mechanisms deliver state-of-the-art results among single-stage probabilistic HMP methods while using less latency and memory than multi-stage diffusion baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2512_00355
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SMamDiff: Spatial Mamba for Stochastic Human Motion Prediction
Fan, Junqiao
Liu, Pengfei
Rao, Haocong
Computer Vision and Pattern Recognition
With intelligent room-side sensing and service robots widely deployed, human motion prediction (HMP) is essential for safe, proactive assistance. However, many existing HMP methods either produce a single, deterministic forecast that ignores uncertainty or rely on probabilistic models that sacrifice kinematic plausibility. Diffusion models improve the accuracy-diversity trade-off but often depend on multi-stage pipelines that are costly for edge deployment. This work focuses on how to ensure spatial-temporal coherence within a single-stage diffusion model for HMP. We introduce SMamDiff, a Spatial Mamba-based Diffusion model with two novel designs: (i) a residual-DCT motion encoding that subtracts the last observed pose before a temporal DCT, reducing the first DC component ($f=0$) dominance and highlighting informative higher-frequency cues so the model learns how joints move rather than where they are; and (ii) a stickman-drawing spatial-mamba module that processes joints in an ordered, joint-by-joint manner, making later joints condition on earlier ones to induce long-range, cross-joint dependencies. On Human3.6M and HumanEva, these coherence mechanisms deliver state-of-the-art results among single-stage probabilistic HMP methods while using less latency and memory than multi-stage diffusion baselines.
title SMamDiff: Spatial Mamba for Stochastic Human Motion Prediction
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.00355