Typical music emotion classification (MEC) approaches categorize emotions and apply pattern recognition methods to train a classifier. However, categorized emotions are too ambigu...
Yi-Hsuan Yang, Yu-Ching Lin, Ya-Fan Su, Homer H. C...
Content-based retrieval has emerged in the face of content explosion as a promising approach to information access. In this paper, we focus on the challenging issue of recognizing ...
Yi-Hsuan Yang, Yu-Ching Lin, Ya-Fan Su, Homer H. C...
The leading voice is an important feature of musical piece and can often be considered as the dominant harmonic source. We propose in this paper a new scheme for the purpose of ef...
Mathieu Lagrange, Luis Gustavo Martins, George Tza...
An audio fingerprint is a compact yet very robust representation of the perceptually relevant parts of audio content. It can be used to identify audio, even when of severely dist...
We present a framework to synchronize pop music to corresponding text lyric. We refine line level alignment achievable by existing work to syllabic level by using a dynamic progra...