TUD |
Institute of Acoustics and Speech Communication |
 |
|
|
Research Signals and systems are models capturing essential properties of the real world. They are fundamental subjects of modern information and communication technology and other research fields. Therefore they play an important role in engineering education. Our chair is responsible for teaching and researching on signals and systems at the Faculty of Electrical an Computer Engineering. The breadth of the field requires focusing on one representative topic. In our case the focus is on signals and systems of speech communication.  The subject of acoutsic pattern recognition is the automatic extraction of information from sound signals. Such signals may, similar to music, be represented by single events (notes or chords) and their temporal sequence (partiture, Demo). Among the tasks of acoustic pattern recognition are the automatic discovery of structures in signals, their mathematical modeling, and the automatic assessment and decision of and upon signals. Applications lie in communication technology, non-destructive testing, machine diagnosis, and in the intelligent processing of musical, bio and speech signals.  Speech and language technology comprises synthesis, coding, transmission and recognition of oral speech as well as automatic text processing. The Institute of Acoustics and Speech Communication has a long research tradition on the field of speech synthesis, starting with electron tube devices up to embedded systems. The classical approaches to speech recognition and synthesis use fundamentally different algorithms and data. Our current reasearch is focused on unifying speech recognition and synthesis using the identical databases and algorithms. Speech and language technology is applied in communication technology, software systems, for multi-modal human-machine interfaces, and for teaching and rehabilitation applications.  Speech and language technology is an important component of modern human-machine interfaces whose deployment and optimization is bound to multiple scenarios. It plays a particularly important role in assisitive techologies which are used to assist blind and otherwise challenged persons. The application of speech and language technology to didactic scenarios - for instance to pronunciation training, language learning, and refinement of instrument playing skills - is an emerging research field. Some important scientific challenges from this field are the automatic assessment of the quality of acoustic signals and the combination of human speech with other communication modes. |
Print