The speech processing group (see: http://www.tik.ee.ethz.ch/~spr) at ETH Zurich is a team of currently six members who work mainly in the domains of speech synthesis and speech recognition. The group is involved in a new interdisciplinary project in the area of multimodal man-machine communication. One of the project aims is to improve the robustness of speech recognition in adverse environment by means of other modalities, mainly vision. From the vision domain, support for the speech recognition can e.g. be attained via lip reading or scene analysis. On the one hand such information can be directly applied in the speech recognizer which is particularly helpful in the presence of strong background noise. On the other hand the information from the visual input may be used on the level of dialog processing to distinguish the user`s speech from the voices of other persons talking nearby.

Applicants must have a master degree in computer science or electrical engineering with a strong background in digital signal processing, speech processing and statistical modeling or related topics.

Please submit your application preferably by email to Dr. Beat Pfister ([email protected]). Your application should contain a CV, a statement of interest, copies of degree certificates and associated scores, the master thesis, and two references.

Email     [email protected]
http://www.tik.ee.ethz.ch/~spr