Title: Automated Transcription of Polyphonic Piano Music A Brief Literature Review
1 Automated Transcription of Polyphonic Piano
MusicA Brief Literature Review
- Catherine Lai
- MUMT-611 MIR
- February 17, 2005
1
2Outline of Presentation
- Introduction
- transcription of polyphonic music
- targeted on specific instruments
- Current state-of-the-art various approaches
- Recent published piano transcription systems
- Dixon, 2000
- Raphael, 2002
- Monti and Sandler, 2002
- Marolt, 2004
- Discussion and Conclusion
- Links to examples of transcription of piano music
recordings - Bibliography
2
3Introduction
- Transcription of polyphonic music
- acoustical waveform --gt parametric representation
- extract pitches, starting times, durations
- First attempt by Moorer, 1975
- note range limitation
- two voices constraint
- Martin, 1996
- piano transcription system up to four voices
- chorale style of J.S. Bach (long durations with
block chords) - Future systems tackled limitations
- targeted system on specific instruments
- Focus of this literature review
- automated transcription of polyphonic piano music
3
4Current State-of-the-Art Various Approaches
- Automated transcription of polyphonic piano music
- input audio files containing polyphonic piano
music - output MIDI representing pitch, timing, volume
- Simon Dixon, 2000. On the Computer Recognition
of Solo Piano Music - standard SP, adaptive peak-picking, pattern
matching - Christopher Raphael, 2002. Automatic
Transcription of Piano Music - HMM
- Monti and Sandler, 2002. Automatic Polyphonic
Piano Note Extraction Using Fussy Logic in a
Blackboard System - blackboard algorithm
- Matija Marolt, 2004 A connectionist approach to
automatic transcription of polyphonic music - neural network models
4
5Published Piano Transcription SystemSimon
Dixon, 2000. On the computer recognition of solo
piano music standardized SP approach
- 1st processing stage
- low-filtering --gt down-sampling signal (12kHz)
- Time-frequency representation
- STFT --gt power spectrum --gt spectral peak
extraction (local maxima gt threshold, adaptive
peak-picking algorithm) - frequency tracks --gt grouping partials --gt
musical notes - Evaluation 13 Mozart piano sonata performed by a
concert pianist - Bösendorfer SE290 computer-monitored piano --gt
MIDI - Results Nno. correctly i.d. notes FPno. note
reported not played FNno. notes played not
reported by system incorrectly I.d. note FP
and FN - score N/(FP FN N)
- recognition accuracy of 70-80
- Future development
- accuracy of dynamic and offset times
5
6Published Piano Transcription SystemChristopher
Raphael, 2002. Automatic transcription of piano
musicHMM
- HMM- trained likelihood model
- statistical pattern recognition and machine
learning for structures - Process
- segment signal to frames extract features
(vector) from frames assign label for content
description - Precise vector features
- total energy (play or silent)
- local burstiness (attack, steady behavior)
- pitch configuration
- Label
- sound pitches collection and re-articulation
(attack, sustain, rest) - Model setup
- hidden process (label process) observable
process (feature vector) - generate reasonable hypotheses for each frame and
construct search graph of the hypotheses
6
7Published Piano Transcription SystemChristopher
Raphael, 2002. Automatic transcription of piano
musicHMM
- Experiment
- Mozart piano sonata
- limitations on range (c two octave below middle c
to the f to two and a half octave above middle c) - number of voices 4 or less
- Evaluation
- borrowed from speech evaluation of Word Error
Recognition Rate - Error Rate 100 (Insertions Deletions
Substitutions) / (Total Words in Truth Sentence) - preliminary results have a Note Error Rate of
39 - 184 substitutions, 241 deletions, 108 insertions
out of 1360 notes - Future improvement
- simple additions may yield better results
- likelihood of chord sequence
- informative note onsets acoustic cues
7
8Published Piano Transcription System Monti and
Sandler 2002. Automatic polyphonic piano note
extraction using fussy logic in a blackboard
system Blackboard algorithm
- Implementation
- Polyphonic Note Recognition using a Fuzzy
Inference System (FIS) as part of the Knowledge
Sources (KSs) in a Blackboard system - Blackboard model arrangement
- hierarchy of data abstraction level
- KSs dictate advancement and is
- activated by Scheduler
- FIS
- take spectral peaks not selected
- create new Note Candidates
- evaluate Candidate by features
- fundamental of note
- harmonic rate
- difference bt max peak in
- spectrum and Candidates fundamental energy
Blackboard system (Monti and Sandler, 2002)
8
9Published Piano Transcription System Monti and
Sandler 2002. Automatic polyphonic piano note
extraction using fussy logic in a blackboard
system Blackboard algorithm
- Evaluation
- 14 piano pieces by various composer including
Beethoven, Mozart, Debussy, Ravel, and Scarlatti - Results Ncorrectly i.d. notes FPnote not
played FNnotes not reported by sys - score N/(FP FN N) Dixons
- detection success rate 45 correct
- 75 correctly detected note / total transcribed
notes
9
10Published Piano Transcription System Matija
Marolt, 2004 A connectionist approach to
automatic transcription of polyphonic piano
music. Neural networks approach
- New model based on networks of adaptive
oscillators was proposed and implemented in SONIC
to partial tracking and note recognition
5.adaptive oscillators try to synchronize to
signals in output freq channels of the auditory
model by adjusting its phase and frequency 6.
When synchronized to the output freq indicate the
freq is periodic and a partial with feq sim to
filter present
1. acoustical waveform --gttime-feq space with an
auditory model 2. auditory model output set of
freq channel 3. periodicity in frequency
channels is related to pitch perception 4. use
adaptive oscillators to calculate periodicity in
frequency channels
76 neural networks others tested multilayer
perception, radial basis function, etc.
Marolt, 2004
10
11Published Piano Transcription System Matija
Marolt, 2004 A connectionist approach to
automatic transcription of polyphonic piano
music. Neural networks approach
- Evaluation
- tested on synthesized and real recordings of
various genre - Results
- synthesized recoding around 90 of all notes
- real recording results not as good (not
available) - most common error (gt 50) octaves and rapidly
played notes (e.g.arpeggios, trills) - greatest challenge very expressive playing
- Chopins Nocturnes
- quiet and almost inaudible left hand
- Further Development
- detecting repeated notes
Marolt, 2004
11
12Discussion and Conclusion
- Various approaches proposed
- standard S.P. techniques HMM blackboard
algorithm neural networks - Common mistakes
- octave, rapid passages, and quiet notes
- Difficulties
- lack standard set of test examples
- evaluation function
- various constraints and formula -- gt comparison
difficult
Piano transcription system Performance results
Dixon 70-80 correct
SONIC 80-95 correct
Raphael 39 wrong
Monti and Sandler 74 correct
12
13Links to examples of transcription of piano music
recordings
- http//lgm.fri.uni-lj.si/matic/SONIC.html
(Marolt) - http//www.ai.univie.ac.at/simon/ (Dixon)
13
14Bibliography
- Dixon, S. 2000. On the Computer Recognition of
Solo Piano Music. Australasian Computer Music
Conference. 31-7. - Marolt, M. 2004. A connectionist approach to
automatic transcription of polyphonic piano
music. IEEE Transactions on Multimedia 6, no. 3
(June) 439-49. - Martin, K. 1996. A blackboard system for
automatic transcription of simple polyphonic
music. MIT Media Laboratory Perceptual Computing
Section Technical Report No. 385. - Montipi, G, and M. Sandler. 2002. Automatic
Polyphonic Piano Note Extraction Using Fuzzy
Logic in a Blackboard System. Proceedings of the
International Conference on Digital Audio
Effects. 39-44. - Moorer, J. 1975. On the segmentation and analysis
of continuous musical sound by digital computer.
Ph.D. thesis, Stanford University, CCRMA. - Raphael, C. 2002. Automatic Transcription of
Piano Music. Proceedings of the International
Conference on Music Information Retrieval.
14