Output details
11 - Computer Science and Informatics
Glasgow Caledonian University
An exploration of the effect of structural and acoustical features on perceived musical emotion
<17> A full paper in the Audio Mostly conference series – a specialist international forum for interdisciplinary cutting edge audio research. Based on EPSRC funded research led by Don Knox (EP/F00558X/1), it describes research utilising a rigorous methodological approach involving musicological expert annotation, and subjective assessment using a large number of participants. The techniques developed mark a significant improvement in music classification model accuracy, which has the potential to impact upon subsequent work on emotion classification. It won the sole best full paper award at the conference, out of around 25 full spoken papers.