Applied Informatics Group

Implementation and Evaluation of Acoustic Distance Measures for Syllables

TitleImplementation and Evaluation of Acoustic Distance Measures for Syllables
Publication TypeMaster Thesis
AuthorsMunier, C.
AbstractIn this work, several acoustic similarity measures for syllables are motivated and successively evaluated. The Mahalanobis distance as local distance measure for a dynamic time warping approach to measure acoustic distances is a measure that is able to discriminate syllables and thus allows for syllable classification with an accuracy that is common to the classification of small acoustic units (60% for a nearest neighbor classification of a set of ten syllables using samples of a single speaker). This measure can be improved using several techniques that however impair the execution speed of the distance measure (usage of more mixture density components for the estimation of covariances from a Gaussian mixture model, usage of fully occupied covariance matrices instead of diagonal covariance matrices). Through experimental evaluation it becomes evident that a decently working syllable segmentation algorithm allowing for accurate syllable border estimations is essential to the correct computation of acoustic distances by the similarity measures developed in this work. Further approaches for similarity measures which are motivated by their usage in timbre classification of music pieces do not show adequate syllable discrimination abilities.
Supervised BySchillingmann, L., and B. Wrede
Academic DepartmentFaculty of Technology
UniversityBielefeld University
Munier2011-MasterThesis.pdf3.02 MB

Recent Best Paper/Poster Awards

Goal Babbling of Acoustic-Articulatory Models with Adaptive Exploration Noise
Philippsen A, Reinhart F, Wrede B (2016)
International Conference on Development and Learning and on Epigenetic Robotics (ICDL-EpiRob) 


Are you talking to me? Improving the robustness of dialogue systems in a multi party HRI scenario by incorporating gaze direction and lip movement of attendees
Richter V, Carlmeyer B, Lier F, Meyer zu Borgsen S, Kummert F, Wachsmuth S, Wrede B (2016)
International Conference on Human-agent Interaction (HAI) 


"Look at Me!": Self-Interruptions as Attention Booster?
Carlmeyer B, Schlangen D, Wrede B (2016)
International Conference on Human Agent Interaction (HAI)


For members