SoundCompass: A Practical Query-by-Humming System

ABSTRACT This paper describes our practical query-by-humming sys- tem, SoundCompass, which is being used as a karaoke song selection system in Japan. First, we describe the fundamen- tal techniques employed by SoundCompass such as normal- ization in a time-wise sense of music data, time-scalable and tone-shiftable time-series data, and making subsequences for efficient matching. Second, we describe techniques to make effective feature vectors based on real music data and do matching with them to develop accurate query-by-humming. Third, we share valuable knowledge that has been obtained through month’s of practical use of SoundCompass. Fourth, we describe the latest version of the SoundCompass system that incorporates these new techniques and knowledge, as well as describe quantitative evaluations that prove the prac- ticality of SoundCompass. The new system provides flexible and accurate similarity retrieval based on k-nearest neighbor searches with multi-dimensional spatial indices structured with multi-dimensional feature vectors.