Learning sound representations using trainable COPE feature extractors

Re-assign community