Xin Jing

My research focuses on guiding AI to better understand people. I investigate how generative models and foundation models can perceive and reason about human affective states, particularly through speech and multimodal communication. By integrating affective understanding into AI systems, my goal is to develop models that go beyond understanding what people say to understanding how they feel, enabling more natural, adaptive, and empathetic human-AI interaction.

Research Interests:

•    Affective Computing: Speech emotion recognition and controllable emotional TTS
•    Multimodal Deep Learning: Aligning language and audio modalities using LLMs, contrastive learning, and diffusion models

Education:

• Modern Communication System (M.Sc.), 2020
• Information Countermeasure Technology (B.Sc.), 2017