M.Sc. degree project proposals in voice science

1. Web-based software ecosystem for speech and voice clinics. This is an overarching topic that includes several exjob-sized subtasks, including

• Browser-based app for recording/editing/playback of patients' voice and speech, with secure upload to existing hospital databases. A fair amount of prior work exists. (audio programming, software engineering, user interface/workflow)
• Automated screening of recording quality, to warn of technical defects, for clinics. A good amount of prior work exists. (signal processing, machine learning)
• Architecture and API for secure archival and retrieval of voice/speech recordings, so that clinics can upload to a regional or national database that can then be accessed by other caregivers and by researchers. (software engineering - security - databases)
• WebAudio implementation of voice/speech analysis by mapping. A great deal of prior work exists, but making it run under WebAudio is a major challenge.

2. Model-based synthesis of singing. This is about re-implementing and developing further the legacy KTH system for synthesis of singing voice, in a modern DSP/music framework. There is a DSP engine component and a separate control component.