Mimicking Audios or Videos

Would be great to have an exercise where you have to mimick a native speakers audio divided by frames and have a feedback taking into account how similar is your audio and the exercise's audio, example: first round divided by sentences, 2nd round a little bit longer and so on. This could be great to improve rhythm, pronunciation and vocabulary.