This thesis studies two problems in music information retrieval: search for a given melody in an audio database, and automatic melody transcription. In both of the problems, the representation of the melody is symbolic, i.e., the melody consists of onset times and pitches of musical notes. In the first part of the thesis we present new algorithms for symbolic melody search. First, we present al...

This paper proposes an algorithm to estimate the melody pitch line (the most dominant pitch sequence) of a given polyphonic audio based on melody range estimation and pitch candidate extraction using a harmonic structure model similar to that proposed by Goto. This paper defines melody pitch candidate as a list of pitch candidates that produces the best-fit harmonic models to the polyphonic aud...

BACKGROUND Transcatheter pulmonary valve implantation using the Melody valve has emerged as an important therapy for the treatment of postoperative right ventricular outflow tract dysfunction. Melody-in-bioprosthetic valves (BPV) is currently considered an off-label indication. We review the combined experience with transcatheter pulmonary valve implantation within BPVs from 8 centers in the Un...

This paper describes the design method of a WWW-based melody-retrieval system which takes a sung melody as a search clue and retrieves the music title from a music database of standard MIDI files(SMF) over the Internet. The most important thing in building a melody-retrieval system on the Internet is to achieve both high matching accuracy and quick search. It was, however, quite difficult to si...

The most dominant source in audio is usually perceived as melody. The estimation of it’s fundamental frequency can be treated as melody estimation. The task become difficult when the polyphony of the audio is high. To determine the most dominant source in human perception, perceptual properties have to be accounted. In this algorithm, we applied psychoacoustic property to the data to estimate t...

the melody analysis of the infant cries was performed manually in the past. based on subjective listening, only estimations could be achieved about the real melodies. a novel method had been introduced a few years ago to categorize the melody shapes. it says that the melodies of the infant cries are combined from elementary units. the melody shapes can be classified according to the order of th...

Two experiments investigated lateral asymmetries in infants' perception of contour-altered and contour-preserved melody changes. In the first study, 40 infants (8.5 months old) of right-handed parents were trained to respond to binaural melody changes with a head turn toward mechanized toy reinforcers. The subsequent test phase included monaural left-ear and right-ear presentations of the famil...

Singing voice is specific in music: a vocal performance conveys both music (melody/pitch) and lyrics (text/phoneme) content. This paper aims at exploiting the advantages of melody and lyric information for real-time audio-to-score alignment of singing voice. First, lyrics are added as a separate observation stream into a template-based hidden semi-Markov model (HSMM), whose observation model is...