Constrain Diarization to specific timestamp of the audio

[KYPcode]

Hello,

Thank you very much for this significant work with NeMo!

I am using the NeMo NFA Aligner to align a transcript to an audio. I then obtain, for each segment of a sentence, a starting timestamp and an ending one. I would like to make them match with the diarization. Is there a way to make the NeMo diarization model to only diarize the specified parts extracted from the NFA in the audio? I tried with the parameters uem_filepath but it is not seems to be took inside the process.