The annotation process is used to describe the embedded information in given audio signals, while the retrieval process
is used to extract an ordered list of audio clips that are relevant to users’ input queries that describe their desired information.