Excellent work. The results are pretty satisfying[0].
I wonder if it could be improved by relying on published lyrics instead of just video transcripts. I say that because the LMFAO song Party Rock Anthem[1] had some awkward moments when the karaoke display showed words that aren't actually part of the song, but are part of the video transcript. I suppose any song that has some sort of scripted intermission/narrative will have this type of problem.
Thank you for trying it out and leaving a comment!
And yes, we are still working on quality improvement. So far it works best on slower song with simpler lyrics and without too much background noise. Having access to the published lyrics will likely enhance the transcription quality for sure. We will look into that if available.
Another feature on our roadmap is too allow users to make corrections. That will likely help in some cases where there are occasional quality issues.
I wonder if it could be improved by relying on published lyrics instead of just video transcripts. I say that because the LMFAO song Party Rock Anthem[1] had some awkward moments when the karaoke display showed words that aren't actually part of the song, but are part of the video transcript. I suppose any song that has some sort of scripted intermission/narrative will have this type of problem.
[0] https://karatube.ai/v/-NuX79Ud8zI
[1] https://karatube.ai/v/KQ6zr6kCPj8