Digital Signal Processing and System Theory

Talk Laila Jassoume

Speech Enhancement for Non-Stationary Distortions (in English)

Date: 22.12.2011, 13:00 h - 14:00 h
Room: Aquarium

Laila Jassoume
CAU, Kiel, Germany

Details (talk will be in English)

If firefighters want to communicate during fire-fighting and rescue operations under respiratory protection there are plenty of technical resources available. Modern breathing masks are equipped with microphones and loudspeakers (within and outside the mask) and also providing a connection to wireless devices is state of the Art. By means of this equipment it is possible for the firefighters to communicate via the radio or directly by recording the speech signal by means of the microphone and reproducing the amplified signal via the integrated loudspeakers. Unfortunately, the speech signal through breath sounds and background noise is greatly distorted.
The signal quality can be significantly improved by the use of digital signal processing. As part of this thesis a speech signal improvement should be realized in real time, which should improve the communication ways:
  • from the microphone to the remote receiver station and
  • from the microphone to the loudspeakers locate on the outside.