You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As mentioned in the paper, Faster RCNN was used for detecting 15 instruments as per AudioSet-instrument dataset but MUSIC dataset contains some other instruments that are not available in MUSIC. How are the extra objects annotated?
The text was updated successfully, but these errors were encountered:
As mentioned in the paper, Faster RCNN was used for detecting 15 instruments as per AudioSet-instrument dataset but MUSIC dataset contains some other instruments that are not available in MUSIC. How are the extra objects annotated?
We did not annotate the object if its category is not in the AudioSet-instrument dataset, hence we exclude this part in the test phase.
As mentioned in the paper, Faster RCNN was used for detecting 15 instruments as per AudioSet-instrument dataset but MUSIC dataset contains some other instruments that are not available in MUSIC. How are the extra objects annotated?
The text was updated successfully, but these errors were encountered: