Application of Deep Learning to IMU sensor motion
Document Type
Conference Proceeding
Publication Date
4-1-2019
Abstract
Deep learning, a sub of machine learning, is a powerful tool for pattern classification but is currently underutilized for IMU motion classification. The digit classification task using the MNIST dataset is one of the most conceptually simple machine learning tutorials and serves as a starting point for other classification tasks. In this paper, we propose to apply deep learning to a set of IMU (inertial measurement unit) sensor-based characters to test how well deep learning works to classify a set of written numbers similar to the open-source MNIST database. Our experiment demonstrates that a deep learning model can correctly classify IMU motion sensor readings tracing out digits in space. These results successfully prepare a deep learning framework for more complex IMU motion classification tasks, such as automatic configuration of grasps and control in biomechatronic prosthetics.
Recommended Citation
Christian, Matthew; Uyanik, Cihan; Erdemir, Erdem; Kaplanoglu, Erkan; Bhattacharya, Sambit; Bailey, Rashad; Kawamura, Kazuhiko; and Hargrove, S. Keith, "Application of Deep Learning to IMU sensor motion" (2019). College of Health, Science, and Technology. 763.
https://digitalcommons.uncfsu.edu/college_health_science_technology/763