A Discriminative Model for Polyphonic Piano Transcription

We present a discriminative model for polyphonic piano transcription. Support vector machines trained on spectral features areused to classify frame-level note instances. The classifier outputs are temporally constrained via hidden Markov models, and the proposed systemis used to transcribe both synthesized and real piano recordings. A frame-level transcription accuracy of 68% was achieved on a newly generated test set, and direct comparisons to previous approaches are provided.