2 files

Technology in Motion Tremor Dataset: TIM-Tremor

posted on 31.01.2019, 00:00 authored by P.J.M. (Paulina) Bank, J. (Jian) Zheng, S.L. (Silvia) Pintea, P.W. (Elma) Ouwehand, S.T. (Suzanne) de Bot, Jan van GemertJan van Gemert, J.J. (Jacobus) van Hilten
Data was collected within the Technology in Motion project (protocol registered as NL54281.058.15), aimed at developing patient friendly and unobtrusive techniques to characterize motor function in patients with neurological disorders. In this context we used video cameras combined with depth sensors (RGB-D sensors, Microsoft KinectTM v2) and 2 accelerometer sensors (ACL300, Biometrics Ltd, Newport, UK) to objectively quantify the frequency and amplitude of tremor of the upper extremity. This dataset contains 55 patient recorded at performing a set of tasks. For each patient and each task, we record a short seqeuence and provide: (i) Kinect RGB video recording, (ii) Kinect depth map recording encoded as a video, (iii) Kinect depth map aligned with the RGB video, (iv) accelerometer recordings of the 2 sensors. We aditionally, provide an overall labeling file containing for each patient: tremor ratings per arm, and tremor diagnosis More detailed description can be found in documentation.



4TU. Centre for Research Data

Time coverage

11 May 2016 – 31 October 2017


media types: application/octet-stream, application/pdf, application/zip, text/csv, text/plain, text/tab-separated-values


Netherlands Organisation for Scientific Research (NWO) Leiden University Medical Center, Department of Neurology, Department of Clinical Neurophysiology TU Delft, Faculty of Electrical Engineering, Mathematics and Computer Science, Computer Vison Lab

Usage metrics