I am trying to map Quranic Verses for the identification of errors. Different audios for the same verse have different durations as it depends on the reciter. I used Audacity to normalize but the duration of verses cannot be normalized. Please can anyone help me prepare the dataset for training such a model? The steps I am following are creating transcripts for each audio verse and then generating its MFCCs. Thanks in advance.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.