Improving Diarisation in Speech-to-Text Applications

In police or legal conversation transcriptions, recognising the people speaking is important and an algorithmically difficult problem. An existing solution is to be extended and improved by additional approaches such as encoder-decoder-attractors or multimodal transcription. Students can actively participate in the definition of the work.

Have we sparked your interest?

I am interested in the study Improving Diarisation in Speech-to-Text Applications and would like to find out more.

Your message Your e-mail Phone number Submit