Speaker: Dr. Björn Krüger
In industrialized societies, a huge percentage of the population suffers from back pain at least once in their lives. The Gokhale Method® addresses back pain by teaching postural modifications. To support posture training, we developed SpineTracker™, a wearable which provides real-time posture feedback to the student.
During Dr. Björn Krüger's talk he will give an overview of the various research activities we do based on SpineTracker’s data. As a first step, we validated the accuracy of our spine shape model by comparing it against motion-capture data in three scenarios: on a test robot, on a moving template, and on the human back. Second, we developed techniques to predict the user’s current position on the basis of the sensor readings. Third, we analyze the student’s stored data and show how patterns in the taught postural changes can be detected and visualized using machine learning techniques. Finally, he’ll present some work on the comparison of sitting styles based on SpineTracker and EMG (electromyography) readings.
As a follow-up, Esther Gokhale, the founder of the Gokhale Method, will introduce the method and demonstrate the SpineTracker wearable in action.
Björn Krüger studied computer science, mathematics, and physics at the University of Bonn. He received his MS in computer science (Dipl.-Inform.) in 2006 and his PhD (Dr. rer. nat.) in computer science in 2012. From 2012 to 2015, he worked as a postdoc at the University of Bonn. In 2015 he joined the Gokhale Method Institute (Stanford, CA) where he leads research activities and the development of the Gokhale SpineTracker™.
Dr. Krüger’s research focuses on techniques to capture, analyze, and synthesize motion data. His doctoral dissertation, Synthesizing Human Motions, focused on data-driven motion synthesis employing large motion-capture collections as its knowledge base. To this end, he developed rapid retrieval methods for huge motion-capture databases, motion recognition and classification techniques, and a new, dynamic motion synthesis procedure. Dr. Krüger also worked as a researcher in the REKOBA (Reconstruction of Movements of Low-Dimensional Sensor and Control Data) project, where human full-body motions were reconstructed based on minimal accelerometer readings. As a postdoc, he was principal researcher at the interdisciplinary GEMMQuad (Generic Motion Models based on Quadrupedal Data) project, funded by the DFG (German Research Foundation). In this project, techniques developed in computer science for human motion data were transferred to the domain of veterinary medicine and biomechanics. During this period, he also developed rapid, unsupervised techniques for semantic-meaningful segmentation of streams of motion data from various sensor modalities, and worked on estimation of soft biometrics from accelerometer readings. Since joining the Gokhale Method Institute, he leads the development of the Gokhale SpineTracker wearable and research activities which study the effectiveness of the Gokhale Method.