Theodoros Kostoulas

Publications (16) View all

• Article: Phone duration modeling: overview of techniques and performance optimization via feature selection in the context of emotional speech

  Alexandros Lazaridis, Todor Ganchev, Theodoros Kostoulas, Iosif Mporas, Nikos Fakotakis
  [show abstract] [hide abstract]
  ABSTRACT: Accurate modeling of prosody is prerequisite for the production of synthetic speech of high quality. Phone duration, as one of the key prosodic parameters, plays an important role for the generation of emotional synthetic speech with natural sounding. In the present work we offer an overview of various phone duration modeling techniques, and consequently evaluate ten models, based on decision trees, linear regression, lazy-learning algorithms and meta-learning algorithms, which over the past decades have been successfully used in various modeling tasks. Furthermore, we study the opportunity for performance optimization by applying two feature selection techniques, the RReliefF and the Correlation-based Feature Selection, on a large set of numerical and nominal linguistic features extracted from text, such as: phonetic, phonologic and morphosyntactic ones, which have been reported successful on the phone and syllable duration modeling task. We investigate the practical usefulness of these phone duration modeling techniques on a Modern Greek emotional speech database, which consists of five categories of emotional speech: anger, fear, joy, neutral, sadness. The experimental results demonstrated that feature selection significantly improves the accuracy of phone duration prediction regardless of the type of machine learning algorithm used for phone duration modeling. Specifically, in four out of the five categories of emotional speech, feature selection contributed to the improvement of the phone duration modeling, when compared to the case without feature selection. The M5p trees based phone duration model was observed to achieve the best phone duration prediction accuracy in terms of RMSE and MAE. KeywordsPhone duration modeling-Statistical modeling-Feature selection-Emotional speech-Text-to-speech synthesis
  International Journal of Speech Technology 04/2012; 13(3):175-188.
• Conference Proceeding: The PlayMancer Database: A Multimodal Affect Database in Support of Research and Development Activities in Serious Game Environment.

  Theodoros Kostoulas, Otilia Kocsis, Todor Ganchev, Fernando Fernández-Aranda, Juan J. Santamaría, Susana Jiménez-Murcia, Maher Ben Moussa, Nadia Magnenat-Thalmann, Nikos Fakotakis
  Proceedings of the International Conference on Language Resources and Evaluation, LREC 2010, 17-23 May 2010, Valletta, Malta; 01/2010
• Conference Proceeding: Affect Recognition in Real Life Scenarios.

  Theodoros Kostoulas, Todor Ganchev, Nikos Fakotakis
  Toward Autonomous, Adaptive, and Context-Aware Multimodal Interfaces. Theoretical and Practical Issues - Third COST 2102 International Training School, Caserta, Italy, March 15-19, 2010, Revised Selected Papers; 01/2010
• Conference Proceeding: Enhancing Emotion Recognition from Speech through Feature Selection.

  Theodoros Kostoulas, Todor Ganchev, Alexandros Lazaridis, Nikos Fakotakis
  Text, Speech and Dialogue, 13th International Conference, TSD 2010, Brno, Czech Republic, September 6-10, 2010. Proceedings; 01/2010
• Source

  Available from: Todor Ganchev

  Conference Proceeding: Feature Selection for Improved Phone Duration Modeling of Greek Emotional Speech.

  Alexandros Lazaridis, Todor Ganchev, Iosif Mporas, Theodoros Kostoulas, Nikos Fakotakis
  Artificial Intelligence: Theories, Models and Applications, 6th Hellenic Conference on AI, SETN 2010, Athens, Greece, May 4-7, 2010. Proceedings; 01/2010