About


Hi, I’m Jaekoo Kang, a researcher and developer specializing in human speech production. Speaking is a fundamental aspect of human communication, where articulators like the tongue and lips shape sounds that are transmitted to the ear and ultimately interpreted by the brain. What fascinates me most is how the continuous movement of articulators is perceived as discrete concepts in our minds—a phenomenon often described as “discrete infinity.”1) This captivating process has fueled my passion for research, and I remain deeply intrigued by the complexities of speech production and its limitless possibilities, especially when combined with computational modeling and AI.



RESEARCH INTERESTS


  • Speech production, articulatory phonology and phonetics
  • Variability and flexibility in skilled articulatory actions
  • Uncontrolled Manifold Framework
  • Articulation-Acoustics Mapping & Inversion
  • Machine learning, deep learning, and computational modeling in speech sciences


TECHICAL INTERESTS


  • Representation of audio and speech (e.g., wav2vec2, WavLM etc.)
  • Forced alignment of speech signals (e.g., HTK, Kaldi, ESPnet etc.)
  • Visualization of speech and audio (e.g., realtime FFT/Spectrogram, Jupyter, Web Audio API, Plotly, P5.js, D3.js etc.)



1) Studdert-Kennedy, M., & Goldstein, L. (2003). Launching Language: The Gestural Origin of Discrete Infinity. In M. Christiansen & S. Kirby (Eds.), Language Evolution (pp. 235–254). Oxford University Press.