EPSRC UnMute project: Developed methods for zero-resource cross-lingual transfer and spoken information retrieval in unwritten languages.
EPSRC COG-MHEAR project: Co-organized the Audio-Visual Speech Enhancement challenge and devised new protocols for evaluating speech enhancement models.
IARPA Material project: Worked on semi-supervised training of ASR models.
H2020 SUMMA project: Developed methods for automatic punctuation prediction in ASR transcripts and was responsible for the deployment of ASR and punctuation models.
CloudASR platform: Developed a cloud-based automatic speech recognition platform supporting both batch and online speech recognition.
MT-ComparEval tool: A tool for comparison and evaluation of machine translation.
Background
Senior Researcher within the ILCC and CSTR institutes of the School of Informatics at the University of Edinburgh. Main research focus is on building speech-to-text systems with limited training data and supervision.