Automated transcription of instrumental music using sheetscore-mapping API and fuzzy classification technique

In the area of digital audio processing, automated music transcription stands as a major enabler in decoding of chord schema so as to facilitate re-arrangement of music. The success rates were found to be largely dependent on the complexity of supplied audio files. Many methodologies can be taken as a base of this purpose of audio transcription, namely genetic transcription algorithm, Wu Pitch Tracking System, Auto-correlation etc. This paper presents an elaborate mapping schema for sheet music transcription which can upon through user acceptance testing be used as an external library for all transcription related research. This paper introduces a new approach to transcription which is simple at its core but requires complete understanding of the nature of sound waves. In this paper fuzzy classification method is used for automated music transcription. Proposed audio processing algorithm takes a monophonic instrumental audio as an input, generates edge detection and stem detection matrices and graphs. The experimental results prove that it gives encouraging results for songs of different tempos.