Automatic Transcription of Speech: From Applied Research to the Market
Speaker: SIEGFRIED JIMMY KUNZMANN, R&D manager of the EML European Media Laboratory GmbH
Date: 15 July 2015 Time: 11:00-13:00
Location: "Stelios Orphanoudakis" Seminar Room, FORTH, Heraklion, Crete
Host: Athanasios Mouchtaris


Over the past few years automatic speech recognition was integrated in several mobile, application scenarios and exposed through powerful smartphones to a huge number of end-users. Algorithmic improvements, higher computing power and access to a large amount of data are key factors for this progress. Nevertheless highly accurate speech recognition in largely varying environments is not yet a “solved problem”. During the talk a brief, high level introduction to the architecture of the EML Transcription Platform will be given and several usage scenarios will be introduced and discussed.


Dr. Siegfried Kunzmann is R&D manager of the EML European Media Laboratory GmbH (Heidelberg, Germany), which is active in the fields of human-computer interaction and scalable, automatic speech transcription. Our transcription platform supports applications ranging from voicemail-to-text, voice messaging & search, speech and media analytics, and broadcast news subtitling to local speech recognition for car, house and media control applications on modern smartphone boards.

He established and headed IBM?s European Voice Technology organization from 1991-2006 which was focusing on the design of multi-lingual speech processing systems (recognition, synthesis, NLU), language technology tools, and research on techniques for language specific needs as well as respective human factors and user interface aspects. He was responsible for the development of all language components offered in IBM?s speech products (with the exception of US English).

Jimmy holds a diploma degree in Computer Science and a PhD in speech processing from the University of Erlangen-Nuremberg. He has published more than 40 papers on various aspects of speech processing technologies including recognition and synthesis algorithms, language specific solutions, and multilingual systems. He is the author of one textbook, has filed more than 10 patents, is member of the International Conference on Text, Speech and Dialogue and was member of the ELRA board.

Conditions of Use | Privacy Policy