Computationally inexpensive and effective scheme for automatic transcription of polyphonic music

We describe a computationally inexpensive but effective scheme for automatic transcription of polyphonic music. Based on a two-step strategy, tracks creation and tracks grouping, the scheme utilizes innovative comb-filtering and ´sharpening´ steps to produce the desired transcription output in the form of discrete notes with temporal, pitch and amplitude attributes. Recall and precision measurements are used to analyze and quantify transcription accuracy. Promising results are achieved for the automatic transcription of both synthetic and acoustic piano music.