A Wearable Motion Capture Dataset for Gait Analysis Using IMUs and Shank-Mounted Egocentric Cameras

We are introducing a novel dataset comprising 14 different walking trials performed by 10 healthy participants with diverse speeds and trial types on overground, treadmill, slope, and stair locomotion modes. This dataset contains a unique combination of inertial measurement unit (IMU) data and shank-mounted egocentric video, synchronized with high-fidelity joint kinematics derived from optical motion capture and musculoskeletal modeling. In total, the dataset provides approximately 327 minutes of synchronized recordings, including IMU signals, motion capture marker trajectories, joint angles, and 588k egocentric video frames per foot. To the best of our knowledge, this is the first dataset offering multimodal recordings of wearable IMU and shank-mounted camera data with ground-truth joint kinematics in a wide range of real-world walking tasks. This dataset is designed to support the development of machine learning models for gait analysis, wearable motion capture, and 3D pose reconstruction in unconstrained environments.