Improving Text to Speech
Is this on the to-do list?
I like the lessons, but la femme robotique doesn't quite deliver the full flavor of the language. I understand the cost of professional voice actors is prohibitive, however, it should be possible, if sentences are reused often enough, to crowd source the voicing similar to how texts are translated. Learners record their pronunciation and vote on the best one. And la femme robotique would always be there to fall back on if a new sentence needs to be generated.
I really don't comprehend the technical challenges of executing such a system. But I feel it would be a fantastic addition to the Duolingo Curriculum, which I feel focuses on the written language as opposed to the spoken. With so much of a languages meaning coming across via intonation, a vital aspect of the learning process is absent.