direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Page Content

[Translate to English:] MagiTact

Lupe
Lupe

[Translate to English:] MagiTact

Hamed Ketabdar

Lupe

Research and development interests:

• General: Statistical Signal Processing, Machine Learning, Artificial Intelligence, Human-Computer Interaction

• Speech Processing: Speech Recognition, Keyword Spotting, Confidence Measurement, Language Modelling, Multi-modal Interaction, Information Retrieval/Extraction/Distillation, Translation, Automatic Annotation

Hamed Ketabdar joined T-Labs as a senior researcher in November 2008.
His research has been mainly on using artificial neural networks for acoustic modelling in speech recognition systems. Before joining the PhD program at Swiss Federal Institute of Technology at Lausanne (EPFL), he did one year of pre-doctoral studies mainly on signal processing and machine learning.

Publications

Journal:

  • Ketabdar, H., Bourlard, H., "Enhancing Phone Posterior Estimation by Integrating Prior and Contextual Knowledge'', submitted to IEEE Transactions on Speech and Audio Processing.
  • Ketabdar, H., Bourlard, H., "Multi-stream Posterior Estimation for Posterior Based Speech Recognition Systems’’, Manuscript under preparation for Speech Communications.

Book Chapter:

  • Hamed Katebdar, Herve Bourlard and Samy Bengio, "Hierarchical Multi-Stream Posterior Based Speech Recognition System'', Machine Learning for Multimodal Interaction (MLMI'05), Lecture Notes in Computer Science, Volume No. 3869, Springler-Verlag.

Selected Conference papers (peer reviewed): For a complete list please check: people.epfl.ch/hamed.ketabdar

  • Hamed Ketabdar, Herve Bourlard, "Hierarchical Integration of Phonetic and Lexical Knowledge in Phone Posterior Estimation”, IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP’08), Las Vegas, USA, April 2008.
  • Hamed Ketabdar, Hynek Hermansky, "Detection of Out-of-Vocabulary Words in Posterior Based ASR'', International Conference on Speech and Language Processing (ICSLP'07), Belgium, August 2007.
  • Hamed Ketabdar, Herve Bourlard, "In-context posteriors as complementary features for TANDEM ASR'', ICSLP'07, Belgium, August 2007.
  • Hamed Ketabdar, Jithendra Vepa, Samy Bengio and Herve Bourlard, "Posterior Based Keyword Spotting with A Priori Thresholds'', ICSLP'06, Pennsylvania, September 2006.
  • Hamed Ketabdar, Jonas Richiardi, and Andrzej Drygajlo, "Global Feature Selection for On-line Signature Verification'', 12th Conference of International Graphonomics Society, pp. 59-63, Salerno, Italy, June 2005.
  • Enhancing Phone Posterior Estimation by Integrating Prior and Contextual Knowledge'', submitted to IEEE Transactions on Speech and Audio Processing.
  • Ketabdar, H., Bourlard, H., "Multi-stream Posterior Estimation for Posterior Based Speech Recognition Systems’’, Manuscript under preparation for Speech Communications.

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions