Research

Speech Sounds and Sociophonetics

Speech sounds, the vowels and consonants, are fundamental elements of speech communication. I study these sounds as complex multimodal structures and try to understand how their acoustic properties (vowel formants, spectral information, etc.,) convey multimodal linguistic information, such as the place of articulation for consonants, and sociolinguistic information, that is information about the dialect of the speaker—a speaker from California or Canada—, the gender, age, etc. The linguistic properties and sociolinguistic variation  are encoded in the speech signal. I combine Machine Learning and experimental approaches to find more information about these issues.

  • Themistocleous, Charalambos (in press). Modern Greek vowels and the nature of acoustic gradience. Phonetica.
  • Themistocleous, Charalambos (2016). The bursts of stops can convey dialectal information. Journal of the Acoustical Society of America. 140(4), EL334-EL340 http://dx.doi.org/10.1121/1.4964818

Prosody: Prosodic variation and change

The melody of speech, also known as prosody, marks constituents in speech as prominent, designates the boundaries of domains (such as the prosodic word and the prosodic phrase), and conveys different melodies (as in questions, statements, and commands). A number of studies have shown that the melody of speech is structured. Understanding this structure and how this structure interacts with the overall cognition and understanding of information is a central objective of my research.

  • Themistocleous, Charalambos (2015). Seeking an anchorage: Evidence from the tonal alignment of the Cypriot Greek prenuclear pitch accent. Language and Speech. Advance on-line publication, doi: 10.1177/0023830915614602. |Link|
  • Themistocleous, Charalambos (2014). Edge-Tone Effects and Prosodic Domain Effects on Final Lengthening. Linguistic Variation. 14(1). 129–160. |Link|
  • Themistocleous, Charalambos (2014). Prosody and Information Structure in Greek (Prosodia kai plirophoriaki domi stin Ellinici). PhD Thesis. University of Athens Greece.

Computational Linguistics and Natural Language Processing

The impressive achievements in areas, such as text-to-speech, speech perception, machine translation, image description generation, and semantic interpretation of the past decade were made possible by employing methods from signal processing, NLP, AI, and machine learning. Methods, such as Deep Neural Networks, SVMs, HMMs, can offer insights about human perception and cogni tion, namely how humans perceive, process, and store information from speech signals, texts, vision, etc., and also how they encode information.  Within this framework, I explore the intersection between speech perception/production and language understanding.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s