Botsa Kishore Kumar supervised by Prof. Suryakanth Gangashetty  received  his doctorate in  Electronics and Communication Engineering (ECE). Here’s a  summary of his research work on  Analysis Detection and Synthesis of High Arousal Speech: 

This thesis investigates the analysis, detection, and synthesis of high-arousal speech, including shouted and Lombard speech commonly produced in emotionally charged or noisy environments. It identifies key acoustic and glottal features that distinguish high-arousal speech from neutral speech and proposes a novel, reference-independent detection method based on intrinsic suprasegmental speech variations. The research further explores the role of pitch (F0), pitch contour, and duration in generating Lombard speech, demonstrating that pitch level has the greatest impact on perceived loudness and intelligibility. Finally, the thesis develops data-driven and transfer learning-based speech synthesis techniques that can generate realistic and intelligible Lombard speech using limited training data. These contributions advance speech processing technologies for robust communication, speech enhancement, and text-to-speech systems operating in noisy environments.

May 2026