Automated Audio Captioning: describing audio content using natural language

Dimitra Emmanouilidou

601 Computer Studies Building

Abstract: This talk is split into two halves: First, I’ll spend some time introducing Microsoft Research, the way we work, and the type of work we do. I will then move on to specific topics of intertest within our research group, Audio and Acoustics; the focus here will be AI for Sound with a quick dive into Automated Audio Captioning, a cross-modal translation task for generating natural descriptions from sound events. 

Bio: Dimitra Emmanouilidou is a Sr Researcher in Microsoft, Redmond, WA, USA. Her interests lie in Signal Processing using Machine Learning and AI approaches, with specific applications in AI for Sound, and in EEG and bio-signal analysis. She serves in the Audio and Acoustic Signal Processing Technical Committee (AASP TC), and has been serving as a reviewer, meta-reviewer, Area and Technical Chair for major conferences and journals in the field. Dimitra received her PhD from the Electrical and Computer Engineering Department at Johns Hopkins University, under Dr. M Elhilali; she holds a M.Sc. in Biomedical Informatics and Technology, under Dr. P Tsakalides, and a B.Sc. in Computer Science from University of Crete, Greece. In her spare time, she enjoys exploring new hiking trails and decorating multi-layered cakes.