New paper “Surface-Free Multi-Stroke Trajectory Reconstruction and Word Recognition Using an IMU-Enhanced Digital Pen”

We are pleased to announce that our latest paper “Surface-Free Multi-Stroke Trajectory Reconstruction and Word Recognition Using an IMU-Enhanced Digital Pen” is now published at Sensors MDPI.

The Paper is published open access: https://www.mdpi.com/1424-8220/22/14/5347

We developed the first end-to-end word trajectory reconstruction system for handwriting using inertial sensors. We trained a Convolutional Neural Network to retrace the movement of a digital pen on paper. Secondly, we trained a Convolutional Recurrent model for handwriting recognition, which achieved comparable results from the generated trajectories relative to the results achieved by the IMU data. Thirdly, we explored modeling the recognition system using both trajectory and IMU data inputs, which showed an improved recognition performance in comparison to the previously trained models.

Congratulations to all authors!

Mohamad Wehbi, Daniel Luge, Tim Hamann, Jens Barth, Peter Kaempf, Dario Zanca and Bjoern Eskofier.