Hi,
I've been using STOI as a metric to compare enhanced audio and clean audio but for a very small number of my test files (~0.1%) I'm getting a negative stoi score. The signals are definitely time-aligned and when listening to these files, it's clear that the intelligibility is very poor and these negative scores are between -0.011 and -0.0043.
Do you know where this issue could be coming from as I'm not sure how to handle it given the score should be between 0-1?
Hi,
I've been using STOI as a metric to compare enhanced audio and clean audio but for a very small number of my test files (~0.1%) I'm getting a negative stoi score. The signals are definitely time-aligned and when listening to these files, it's clear that the intelligibility is very poor and these negative scores are between -0.011 and -0.0043.
Do you know where this issue could be coming from as I'm not sure how to handle it given the score should be between 0-1?