Manually annotate audio samples to find the important segments
Closed this issue · 2 comments
mwinton commented
Rather than assuming that the entire audio file contains a "positive" example of a class, I think we may need to manually annotate the appropriate portions.
mwinton commented
It looks like this won't be required for the majority of classes, but leaving this open for now until we have a chance to look into the 11 classes with poor accuracy.
paloukari commented
Closing