Shape-aware Inertial Poser: Motion Tracking for Humans with Diverse Shapes Using Sparse Inertial Sensors

2025-10-19

AI Summary

Existing sparse-IMU motion capture methods heavily rely on template-based adult body models, limiting generalization to populations with significant anatomical variation—e.g., children. This work is the first to systematically model how body shape affects IMU observations, proposing a body-shape-aware inertial pose estimation framework. We design an MLP-based, shape-conditioned network to decouple motion across diverse morphologies; introduce a regression-based acceleration bias correction module that accounts for shape-induced inertial deviations; and integrate physics-informed optimization with joint velocity mapping to reconstruct full-body motion from only sparse IMUs. Evaluated on the first cross-body-type benchmark dataset—comprising 10 children and 10 adults (heights 110–190 cm, 400 minutes total)—our method achieves substantial gains in generalization accuracy. Code and data are publicly released.

Technology Category

• Computer Vision: Motion & Tracking
• Intelligent Robots: State Estimation
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Security and Privacy: Large-scale security measurements
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Human motion capture with sparse inertial sensors has gained significant attention recently. However, existing methods almost exclusively rely on a template adult body shape to model the training data, which poses challenges when generalizing to individuals with largely different body shapes (such as a child). This is primarily due to the variation in IMU-measured acceleration caused by changes in body shape. To fill this gap, we propose Shape-aware Inertial Poser (SAIP), the first solution considering body shape differences in sparse inertial-based motion capture. Specifically, we decompose the sensor measurements related to shape and pose in order to effectively model their joint correlations. Firstly, we train a regression model to transfer the IMU-measured accelerations of a real body to match the template adult body model, compensating for the shape-related sensor measurements. Then, we can easily follow the state-of-the-art methods to estimate the full body motions of the template-shaped body. Finally, we utilize a second regression model to map the joint velocities back to the real body, combined with a shape-aware physical optimization strategy to calculate global motions on the subject. Furthermore, our method relies on body shape awareness, introducing the first inertial shape estimation scheme. This is accomplished by modeling the shape-conditioned IMU-pose correlation using an MLP-based network. To validate the effectiveness of SAIP, we also present the first IMU motion capture dataset containing individuals of different body sizes. This dataset features 10 children and 10 adults, with heights ranging from 110 cm to 190 cm, and a total of 400 minutes of paired IMU-Motion samples. Extensive experimental results demonstrate that SAIP can effectively handle motion capture tasks for diverse body shapes. The code and dataset are available at https://github.com/yinlu5942/SAIP.

Problem

• Tracking human motion with sparse inertial sensors across diverse body shapes
• Addressing body shape variations affecting IMU-measured acceleration data
• Developing shape-aware inertial motion capture for non-template body types

Innovation

• Decomposes sensor measurements into shape and pose components
• Transfers IMU accelerations to template body via regression model
• Maps joint velocities back to real body with shape optimization