A Sound-Retrieval Index Based on Two-Dimensional Similarity Maps
Treatment of timbre in computer music systems is an unsolved problem. A practical aspect is the organization of sound material in a sound data base. Our approach uses a connectionist model to transform multidimensional timbre on a two-dimensional similarity index for sounds. The authors propose an algorithm that simulates essential properties of the peripheral auditory system and realizes a self-organizing mapping of an input sound space to a representation of lower dimension.
This paper costs $20 for non-members, $5 for AES members and is free for E-Library subscribers.