Humming transcription Our frontend for music search Goal Convert humming into an MIDI note sequence Steps involved in humming transcription Track pitch over the humming input Identify notes via silence ID: 309875
Download Presentation The PPT/PDF document "Humming Transcription" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Humming Transcription
Humming transcription
Our front-end for music search
Goal: Convert humming into an MIDI note sequence
Steps involved in humming transcription
Track pitch over the humming input
Identify notes via silence
Find the best MIDI sequenc
e (of integer semitones) via key transpositionSlide2
Pitch Tracking over the Humming
Pitch tracking methods
ACF-based peak picking
ACF-based dynamic programmingExample:Humming inputIdentified pitch vectorNotes are identified by volume or clarity thresholding Slide3
MIDI Sequence Identification
We need to represent each note as an integer semitone (or the so-called MIDI number)
Perform exhaustive search on pitch shift to identify the best integer semitones
ExampleNotes obtained as medians from
pitch vector: [50.56, 49.45, 56.46, 56.36, 58.82, 58.65, 56.70, 54.74, 54.93, 53.85, 53.70, 51.53, 51.73, 49.62]Best integer notes after key transposition: [50, 50, 57, 57, 59, 59, 57, 55, 55, 54, 54, 52, 52, 50]Slide4
Humming Transcription
Humming transcription is our front-end for
music search.
The process for humming transcriptionTypical hummingPitch vector obtained from
ACF: [0, 0, 49.5336, 49.4312, 49.4817, 50.0910, 0, 50.5623, ...]Median notes obtained from pitch vector: [50.5623, 49.4559, 56.4628, 56.3636, 58.8217, 58.6510, 56.7025, 54.7415, 54.9399, 53.8507, 53.7002, 51.5369, 51.7357, 49.6283]
Best integer notes after key transposition: [
50, 50, 57, 57, 59, 59, 57, 55, 55, 54, 54, 52, 52, 50]