Music transcription consists in transforming the musical content of audio data in to a symbolic representation. The objective of this study is to investigate a transcription system for polyphonic piano, triggered by events corresponding to the played notes. The proposed methodfocuses on note events and their main characteristics: the attack instant, the pitch and the final instant. Onset detection exploits a binary time-frequency representation of the audio signal. Note classification and offset detection are based on constant Q transform (CQT) and support vector machines (SVMs). We present a collection of experiments using synthesized MIDI files and piano recordings, and compare the results with existing approaches.
Event based transcription system for polyphonic piano music
Giovanni Costantini;
2009
Abstract
Music transcription consists in transforming the musical content of audio data in to a symbolic representation. The objective of this study is to investigate a transcription system for polyphonic piano, triggered by events corresponding to the played notes. The proposed methodfocuses on note events and their main characteristics: the attack instant, the pitch and the final instant. Onset detection exploits a binary time-frequency representation of the audio signal. Note classification and offset detection are based on constant Q transform (CQT) and support vector machines (SVMs). We present a collection of experiments using synthesized MIDI files and piano recordings, and compare the results with existing approaches.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.