Sound characteristics (features) are taken per e.g. 20 ms, but manual annotation takes place in durations of seconds. If we want to use these characteristics along with labels, how can this be performed? is there some form of grouping?
Add a comment
|