Privacy-Preserving Musical Database Matching

In this paper we present an illustratory process which allows privacy-preserving transactions in the context of musical databases. In particular we address the problem of matching a piece of music audio to a service database in such a way such that the database provider will not directly observe the query, nor its result, thereby preserving the privacy of the inquirer. We formulate this process within the field of secure multiparty computation and show how such a transaction can be achieved once we derive secure versions of basic signal processing operations.