Welcome!

My name is Seong-Gyun Leem. I received my Ph.D. in Electrical Engineering at The University of Texas at Dallas. My research focuses on developing advanced speech and multimodal signal processing systems for real-world applications, considering noise robustness, computational efficiency, user privacy, and controllability in machine learning models.

Research interests

Noise-robust speech emotion recognition

Efficient deep learning for real-world applications

Controllable speech/audio synthesis model

My main research goal is to enable speech-based emotional communication between machines and humans in real-world applications. I aim to solve practical problems of applying spoken language and behavioral signal processing systems to real-world applications. To achieve this, I work on developing noise-robust emotion recognition, efficient domain adaptation, and personalization algorithms for speech and behavioral signal processing systems. My research interests include multimodal emotion recognition, multilingual speech processing, data augmentation, and model compression, which can enhance the applicability of speech and behavioral signal processing systems in real-world scenarios.