George P. Kafentzis

George P. Kafentzis is a Lecturer in the Computer Science Department at the University of Crete, where he teaches Physics for Engineers (CS-112), Digital Signal Processing (CS-370), and Signals and Systems (CS-215). He is a core member of the Speech Signal Processing Lab within the Multimedia Informatics Labs, focusing on advanced signal processing methodologies.

His educational background includes a Ph.D. in Signal Processing and Telecommunications from MATISSE Doctoral School (University of Rennes 1) and a Ph.D. in Computer Science and Engineering from the University of Crete (2014), a Master of Science in Computer Science (2010), and a Bachelor's degree in Computer Science (2008), all from the University of Crete.

Research interests span speech, audio, and biosignal processing with emphasis on sinusoidal modeling, emotion recognition from speech, deep learning applications, pathological speech analysis, and music signal processing. His work bridges theoretical signal processing with clinical and engineering applications, particularly in non-invasive vocal fold pathology detection through glottal analysis. Recent publications demonstrate a strategic pivot toward cough sound analysis for respiratory diagnostics using AI, while maintaining core expertise in adaptive sinusoidal models for speech transformations.

Publication trends reveal an evolution from fundamental speech modeling (2010-2016) toward applied health informatics (2021-present), with increasing focus on real-world diagnostic systems leveraging cough acoustics. Over 50% of recent work integrates deep learning with traditional signal processing for medical applications, particularly in low-resource settings.

• Graduate student Scholarship - Institute of Computer Science, FO.R.T.H. (2008-2010)
• Undergraduate Scholarship - Institute of Computer Science, FO.R.T.H. (2007-2008)

As an active industry collaborator, Kafentzis has served as Signal Processing Engineer at Hyfe AI (2022-2025) and contractor for VoiceSignals and Toshiba Research Europe. His teaching portfolio includes a widely adopted textbook Continuous and Discrete Time Signal Processing (2019), which integrates MATLAB implementations with theoretical foundations. Current research leverages his signal processing expertise in cough monitoring systems validated through multicenter clinical trials.

He leads projects in the Speech Signal Processing Lab including Novel Deep Learning Architectures for Automatic Speech Recognition and Speech Emotion Recognition and Visualization Techniques, with recent work extending to Greek-language pathological speech analysis and respiratory health monitoring systems.