I'm looking for an API or model or a python code which could convert YouTube videos with disabled transcripts into text keeping the intent of academic content. Also, it should take Indian Accent as well, as an input and languages to be considered are majorly English, Hindi and Marathi.
I tried with YouTube's API but it converts only those videos which has their transcripts enabled, also, I tried the AssemblyAI's APIs which was converting videos into text for even those videos whose transcripts are disabled but the accuracy was very low, and, it is not taking Indian accent properly.I want an API or model/multi-model which could get the transcripts of even those YouTube videos which has disabled transcripts with higher accuracy and languages preferred are English, Hindi and Marathi.