Thomas Winkler

I am a researcher in the speech group in NetMedia at Fraunhofer IAIS. In 2005, I received my diploma in Electrical Engineering from RWTH Aachen University, and I started to work for Fraunhofer in 2006. During my university studies I focused on multimedia signal processing and classification - in particular approaches for blind source seperation. My current work at Fraunhofer includes research of various aspects of multimedia pattern recognition from audio-visual similarity to robust automatic speech recognition.

Research Interests

  • speech analysis
  • automatic speech recognition
  • noise robust speech processing
  • multimedia pattern recognition
  • speech corpora
  • topic and genre detection
  • semantic media analysis

Miscellaneous

Co-organiser of the workshop EVENTS2010.

Guest Editor of Special Issue on Event Recognition in Applied Artificial Intelligence AAI journal.

Publications

  • Winkler, T.: How Realistic is Artificially Added Noise? Proceedings of the 12th Annual Conference of the International Speech Communication Association ISCA (INTERSPEECH), 2011
  • Baum, D.; Schneider, D.; Bardeli, R.; Schwenninger, J., Samlowski, B.; Winkler, T. & Köhler, J.: DiSCo - A German Evaluation Corpus for Challenging Problems in the Broadcast Domain. Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10), 2010
  • Winkler, T. & Bardeli, R.: An integrated approach of a robust command and control application on the motorcycle. Proceedings of the 13th International Conference SPEECH and COMPUTER: SPECOM 2009, 2009
  • Winkler, T.; Pronkine, S.; Bardeli, R. & Köhler, J.: A study of throat microphone performance in automatic speech recognition on motorcycles. NAG/DAGA 2009 International Conference on Acoustics, 2009
  • Baum, D.; Samlowski, B.; Winkler, T.; Bardeli, R. & Schneider, D.: DiSCo - a speaker and speech recognition evaluation corpus for challenging problems in the broadcast domain. Proceedings of the GSCL Symposium "Sprachtechnologie und eHumanities", 2009
  • Konstantopoulos, S.; Pottebaum, J.; Schon, J.; Schneider, D.; Winkler, T.; Paliouras, G. & Koch, R.: Ontology-Based Rescue Operation Management. Mobile Response: Second International Workshop on Mobile Information Technology for Emergency Response, MobileResponse 2008, Springer, 2009
  • Winkler, T.; Kostoulas, T.; Adderley, R.; Bonkowski, C.; Ganchev, T.; Köhler, J. & Fakotakis, N.: The MoveOn Motorcycle Speech Corpus. Proceedings of the Sixth International Language Resources and Evaluation (LREC'08), 2008
  • Schneider, D.; Winkler, T.; Löffler, J. & Schon, J.: Robust Audio Indexing and Keyword Retrieval Optimized for the Rescue Operation Domain. Mobile Response: First International Workshop on Mobile Information Technology for Emergency Response, MobileResponse 2007, Springer, 2007

Latest blog entries

Our Special Issue on Event Recognition of the Taylor & Francis Journal series on Applied Artificial Intelligence is available online now.

Categories

I will visit the ASRU Workshop 2011 on Big Island, Hawaii, in a bit more than one week. I am already excited about the scientific exchange during this first class workshop. But I am even more excited, that my colleague Daniel and I have the chance to present two demos at the "Show & Tell" session of the workshop. The demos show two rather different fields of application for automatic speech recognition for video websites: video citation and contextual video advertising.

Categories

Last Tuesday the first Targeted Advertising Congress ("Targeted Advertising - Forschung & Praxis") took place in the Kameha Grand Bonn. About 90 representatives from companies and research institutions met to discuss challenges, innovations and research in the area of targeted advertising.

Categories