Blog

Research

Our research focuses on audio intelligence and media systems, with active work in historical music restoration and speech emotion recognition.

2 min read

Speech Emotion Recognition

Models that infer emotional and affective states from speech, with attention to generalization, uncertainty, and real-world conditions.

Read note

Historical Music Restoration

Machine-learning and signal-processing methods for restoring historical music recordings while preserving their musical and historical character.