Signal processing techniques, and information technology in general, have undergone several scientific advances which permit us to address the very complex problem of automatic music transcription (AMT). During the last ten years, the interest in AMT has increased rapidly, and the time has come for a book-length overview of this subject. The purpose of this book is to present signal processing algorithms dedicated to the various aspects of music transcription. AMT is a multifaceted problem, comprising several subtasks: rhythm analysis, multiple fundamental frequency analysis, sound source separation, musical instrument classification, and integration of all these into entire systems. AMT is, in addition, deeply rooted in fundamental signal processing, which this book also covers. As the field is quite wide, we have focused mainly on signal processing methods and Western polyphonic music. An extensive presentation of the work in musicology and music perception is beyond the scope of this book.