🎉 Two Papers Newly Accepted for UbiComp 2025 Presentation!

Overview

We are pleased to share that two of our papers from the IMWUT February cycle have been accepted and will be presented at UbiComp 2025: Motion2Press and KineticsSense. Together, these projects explore how wearable and mobile sensing can move beyond coarse activity recognition toward richer biomechanical understanding of human movement in everyday settings.

Both papers focus on lower-limb motion analysis, but from complementary perspectives. Motion2Press studies how lightweight IMU sensing can infer plantar pressure information for gait analysis, while KineticsSense integrates multimodal wearable signals to model lower-limb motion kinetics and muscle activation patterns. We are excited to share these works with the ubiquitous computing community and discuss their implications for rehabilitation, sports science, health monitoring, and future wearable intelligence.

Accepted Papers

Motion2Press

Motion2Press: Cross Model Learning from IMU to Plantar Pressure for Gait Analysis proposes a cross-modal learning framework that uses inertial measurement units (IMUs) to infer plantar pressure distribution, ground reaction force, and center of pressure. Plantar pressure is highly informative for gait analysis, clinical assessment, and sports training, but traditional measurement systems are often expensive, constrained to laboratory settings, and difficult to deploy in daily life.

Motion2Press addresses this gap by learning from IMU signals to reconstruct pressure-related information in a more lightweight and practical way. The work contributes a data-driven pipeline for estimating gait-relevant biomechanical signals using minimal wearable sensing, supporting more accessible motion analysis outside specialized labs.

Publication page: Motion2Press

KineticsSense

KineticsSense: A Multimodal Wearable Sensor Framework for Modeling Lower-Limb Motion Kinetics investigates how wearable sensing can capture not only movement kinematics, but also the underlying kinetics of human motion. Current motion analysis often focuses on visible movement patterns, while biomechanical factors such as force generation and muscle activation remain harder to access in real-world environments.

KineticsSense combines IMU and plantar pressure data to estimate lower-limb electromyography (EMG) signals, enabling a richer representation of human movement. The system is evaluated across activities such as walking, running, squats, and jumps, with case studies highlighting its potential for rehabilitation assessment and athletic performance analysis.

Publication page: KineticsSense

Looking Ahead

We look forward to presenting these papers at UbiComp 2025 and exchanging ideas with researchers working on wearable sensing, mobile health, biomechanics, and human-centered AI. More presentation details will be added as the conference approaches.