In developing automated systems to recognize the emotional content of music, we are faced with a problem spanning two disparate domains: the space of human emotions and the acoust...
Erik M. Schmidt, Douglas Turnbull, Youngmoo E. Kim
Emotional expression when performing music (singing or playing musical instruments) requires skill, but such a skill is generally difficult to learn. Computer systems that can mak...
A simple note onset detection system for music is presented in this work. To detect onsets, a 1/12 octave filterbank is simulated in the frequency domain and the band derivatives ...
Music classification continues to be an important component of music information retrieval research. An underutilized tool for improving the performance of classifiers is feature ...
Abstract—We present the query-by-description (QBD) component of “Kandem,” a time-aware music retrieval system. The QBD system we describe learns a relation between descriptiv...