About This Topic
Implement a smartglass application that converts a speech audio input stream into poetry by reading lip movements and converting them into proper TTS read poetry provided by a database. One application could be to overlay a blubbering partner with poems of Walt Witman; another to overlay a ranting boss by Schiller, etc. Requires good knowledge in computer vision and audio retrieval (ideally, courses Similarity Modeling 1/2) as well as minimal Android programming knowledge. If done as a diploma thesis, it has to include an extended user study based on real smart glasses (e.g. from Vuzix or Toshiba).