Sir, what are the parameters can be used to evaluate sphinx3_align results. To find root mean square error between ground truth and obtained results is not possible because the ground truth is not available. To manually label the boundaries is also not possible because I need to evaluate for large utterance. Please help me in this regard. Thank you.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Sir, what are the parameters can be used to evaluate sphinx3_align results. To find root mean square error between ground truth and obtained results is not possible because the ground truth is not available. To manually label the boundaries is also not possible because I need to evaluate for large utterance. Please help me in this regard. Thank you.
I have searched so many
Papers everybody is concentrating on decoder results, how to assess the aligned results. Please tell me. Thank you.
Create ground truth manually aligning labels.
There are datasets with manually assigned phone boundaries, for example TIMIT. You can use it for evaluation.
Ok, sir thanks for your quick reply and suggestion