Efficient and Simple Algorithms for Time-Scaled and Time-Warped Music Search

We present new algorithms for time-scaled and time-warped search from symbolic polyphonic music database. Our algorithm for timescaled search works in O(nm) time and O(m) space, where n denotes the number of notes in the database and m denotes the number of notes in the pattern. Our algorithm for time-warped search works in O(n(m+ logn)) time and O(n) space. If the set of possible pitch values is constant, the time complexity of the latter algorithm is only O(nm). The new algorithms are more efficient than the earlier algorithms, both in theory and practice. Moreover, the new algorithms are conceptually simple.