Audio Signal Classification: An Overview

Audio signal classification consists of extracting physical and perceptual features from a sound, and of using these features to identify into which of a set of classes the sound is most likely to fit. The feature extraction and classification algorithms used can be quite diverse depending on the classification domain of the application. This paper presents an overview of the current state of the audio signal classification research literature. 1 Introduction Audio signal classification (ASC) is a field of research that has historically been explored in a few very concentrated areas, with less work done on the general problem. The individual pieces have traditionally been speech recognition and related problems, music transcription [Moo77] [Pis86], and recently speech/music discrimination [Sau96] [SS97]. Other problems in the field have been researched as well, but the main direction has been toward speech and music applications. The major exception that has recently appeared is the..