VersaPants: A Loose-Fitting Textile Capacitive Sensing System for Lower-Body Motion Capture

TL;DR

VersaPants tackles the challenge of unobtrusive, lower-body motion capture by embedding capacitive textile sensors in loose-fitting pants and performing on-device pose estimation with a lightweight Transformer-based model. The system integrates a VersaSens-based electronic backbone, a capacitive sensing board with 16 channels, and a compact CNN2D+Transformer that regresses four lower-body joints in the SMPL space from 4-second textile windows, enabling real-time inference on a smartwatch without quantization. Evaluation across 11 participants and 16 movements shows competitive MPJPE and MPJAE, with favorable edge-deployment metrics (≈42 FPS on a consumer smartwatch) and strong generalization under LOPO and LOEO protocols. The work demonstrates a practical path toward comfortable, privacy-preserving, embedded motion capture for fitness, healthcare, and wellbeing, while outlining key limitations in ground-truth fidelity, hardware robustness, and dataset diversity for future improvements.

Abstract

We present VersaPants, the first loose-fitting, textile-based capacitive sensing system for lower-body motion capture, built on the open-hardware VersaSens platform. By integrating conductive textile patches and a compact acquisition unit into a pair of pants, the system reconstructs lower-body pose without compromising comfort. Unlike IMU-based systems that require user-specific fitting or camera-based methods that compromise privacy, our approach operates without fitting adjustments and preserves user privacy. VersaPants is a custom-designed smart garment featuring 6 capacitive channels per leg. We employ a lightweight Transformer-based deep learning model that maps capacitance signals to joint angles, enabling embedded implementation on edge platforms. To test our system, we collected approximately 3.7 hours of motion data from 11 participants performing 16 daily and exercise-based movements. The model achieves a mean per-joint position error (MPJPE) of 11.96 cm and a mean per-joint angle error (MPJAE) of 12.3 degrees across the hip, knee, and ankle joints, indicating the model's ability to generalize to unseen users and movements. A comparative analysis of existing textile-based deep learning architectures reveals that our model achieves competitive reconstruction performance with up to 22 times fewer parameters and 18 times fewer FLOPs, enabling real-time inference at 42 FPS on a commercial smartwatch without quantization. These results position VersaPants as a promising step toward scalable, comfortable, and embedded motion-capture solutions for fitness, healthcare, and wellbeing applications.

Figures (8)

• Figure 1: Design of Data Acquisition Unit: a) Capacitive Sensing Boards with 4 FDC2214 units, supporting up to 16 channels. b) 3D CAD design of the plastic enclosure. c) The final DAU with manufactured PCBs and printed casing.
• Figure 2: Production of VersaPants: a)Sensor placements for the pants. b) Layered construction of each channel. c) Side view of the pants with wires channeled through the inner seams. d) Frontal view of the pants. e) Rear view of the pants, with DAU attached to the back part of the waist.
• Figure 3: Deep Learning Pipeline for the Proposed CNN2D+Transformer Architecture.
• Figure 4: Results of Leave-One-Participant-Out evaluation with our proposed CNN2D+Transformer architecture.
• Figure 5: MPJPE and MPJAE Metrics Averaged Per Joint. Error bars represent the interparticipant standard deviation for joint errors.