Artifact Removal Algorithm for an EMG-based Silent Speech Interface
by , , , ,
Abstract:
An electromygraphic (EMG) Silent Speech Interface is a system which recognizes speech by capturing the electric potentials of the human articulatory muscles, thus enabling the user to communicate silently. This study deals with improving the EMG signal quality by removing artifacts: The EMG signals are captured by electrode arrays with multiple measuring points. On the resulting high-dimensional signal, Independent Component Analysis is performed, and artifact components are automatically detected and removed. This method reduces the Word Error Rate of the silent speech recognizer by 9.9% relative on a development corpus, and by 13.9% relative on an evaluation corpus.
Reference:
Artifact Removal Algorithm for an EMG-based Silent Speech Interface (Michael Wand, Adam Himmelsbach, Till Heistermann, Matthias Janke, Tanja Schultz), In International Conference of the IEEE Engineering in Medicine and Biology Society, Osaka, Japan, 2013. (EMBC 2013)
Bibtex Entry:
@inproceedings{wand2013artifact,
  year={2013},
  title={Artifact Removal Algorithm for an EMG-based Silent Speech Interface},
  note={EMBC 2013},
  booktitle={International Conference of the IEEE Engineering in Medicine and Biology Society, Osaka, Japan},
  url={https://www.csl.uni-bremen.de/cms/images/documents/publications/WandSchultz_EMBC2013.pdf},
  abstract={An electromygraphic (EMG) Silent Speech Interface is a system which recognizes speech by capturing the electric potentials of the human articulatory muscles, thus enabling the user to communicate silently. This study deals with improving the EMG signal quality by removing artifacts: The EMG signals are captured by electrode arrays with multiple measuring points. On the resulting high-dimensional signal, Independent Component Analysis is performed, and artifact components are automatically detected and removed. This method reduces the Word Error Rate of the silent speech recognizer by 9.9% relative on a development corpus, and by 13.9% relative on an evaluation corpus.},
  author={Wand, Michael and Himmelsbach, Adam and Heistermann, Till and Janke, Matthias and Schultz, Tanja}
}