In recent years, neural networks have been used to generate music pieces, especially symbolic melody. However, the long-term structure in the melody has posed great difficulty for designing a good model. In this paper, we present a hierarchical recurrent neural network for melody generation, which consists of three Long-Short-Term-Memory (LSTM) subnetworks working in a coarse-to-fine manner. Sp...

This paper describes a method for melody detection in polyphonic musical signals. Our approach starts by obtaining a set of pitch candidates for each time frame, with recourse to an auditory model. Trajectories of the most salient pitches are then constructed. Next, note candidates are obtained by trajectory segmentation (in terms of frequency and pitch salience variations). Too short, low-sali...

We submit a method for the MIREX 2008 task “Audio Melody Extraction”. The method first transcribes input audio file into MIDI notes and then estimates fundamental frequency (F0) trajectory for each note. The melody transcription method uses a frame-wise pitch salience estimator to measure the strength of different fundamental frequencies. HMMs are used to model melody notes, other instrument no...

This paper is about the use of perceptual principles for melody estimation. The melody stream is understood as generated by the most dominant source. Since the source with the strongest energy may not be perceptually the most dominant one, it is proposed to study the perceptual properties for melody estimation: loudness, masking effect and timbre similarity. The related criteria are integrated ...

As reported in the recent literature on piano performance, an emphasized voice (the melody) tends to be played not only louder than the other voices, but also about 30 ms earlier (melody lead). It remains unclear whether pianists deliberately apply melody lead to separate different voices, or whether it occurs because the melody is played louder (velocity artifact). The velocity artifact explan...

Infants 8-11 months of age were exposed to repetitions of a 6-tone sequence or melody, then tested for their discrimination of transpositions of that sequence as well as other melodic transformations previously used by Massaro, Kallman, and Kelly with adults. In experiment 1, infants showed evidence of discriminating all transformations from the original melody. In Experiment 2, the task was ma...

In two experiments, the perceptual similarity between a strong tonal melody and various transpositions was investigated using a paradigm in which listeners compared the perceptual similarity of a melody and its transposition with that of the same melody and another transposition. The paradigm has the advantage that it provides a direct judgment regarding the similarity of transposed melodies. T...

This paper describes an efficient method for the identification of the melody voice from the frame-wise updated magnitude and frequency values of tone objects. Most state of the art algorithms employ a probabilistic framework to find the best succession of melody tones. Often such methods fail, if there are several musical voices with a comparable strength in the audio mixture. In this paper, w...

We introduce a method for theMIREX 2006 “AudioMelody Extraction” task in which the goal is to estimate fundamental frequency (F0) trajectory of the main melody within polyphonic music. The introduced method is based on multipleF0 estimation followed by acoustic and musicological modeling. The acoustic model consists of separate models for melody notes and for no-melody segments. The musicologic...

Adults and 9- to 11-month-old infants listened to repeated transpositions of either a prototypical Western melody that was based on the major triad or a nonprototypical Western melody that was based on the augmented triad. In both cases, the transpositions were either to related keys (standing in a 2:3 frequency ratio) or to unrelated keys (more complex frequency ratios). For the prototypical m...