MIR workshop 2011

From CCRMA Wiki

(Difference between revisions)
Jump to: navigation, search
(Lectures & Labs)
(Logistics)
 
(45 intermediate revisions not shown)
Line 8: Line 8:
  - Jay LeBoeuf, [http://www.imagine-research.com Imagine Research ]
  - Jay LeBoeuf, [http://www.imagine-research.com Imagine Research ]
  - Rebecca Fiebrink, [http://www.cs.princeton.edu/~fiebrink/Rebecca_Fiebrink/welcome.html Princeton University]
  - Rebecca Fiebrink, [http://www.cs.princeton.edu/~fiebrink/Rebecca_Fiebrink/welcome.html Princeton University]
-
  - Doug Eck, Google [http://www.google.com Google]
+
  - Douglas Eck, Google Research [http://research.google.com Google]
  - Stephen Pope,  [http://www.imagine-research.com Imagine Research ]
  - Stephen Pope,  [http://www.imagine-research.com Imagine Research ]
  - Steve Tjoa, University of Maryland / [http://www.imagine-research.com Imagine Research ]
  - Steve Tjoa, University of Maryland / [http://www.imagine-research.com Imagine Research ]
  - Leigh Smith,  [http://www.imagine-research.com Imagine Research ]
  - Leigh Smith,  [http://www.imagine-research.com Imagine Research ]
-
  - George Tzanetakis, [http://webhome.cs.uvic.ca/~gtzan/ University of Victoria]
+
  - George Tzanetakis, Google Research, [http://webhome.cs.uvic.ca/~gtzan/ University of Victoria]
* Participants:
* Participants:
 +
- Chris Colatos, [http://jeffalbert.com Jeff Albert], Kamlesh Lakshminarayanan, Sean Zhang, Eli Stine, David Bird, Gina Collecchia, Dekun Zou, Bill Paseman, John Amuedo.
 +
- See photo here: [http://stevetjoa.com/568 CCRMA MIR Workshop 2011 Participants]
== Abstract ==  
== Abstract ==  
Line 29: Line 31:
== Lectures & Labs ==
== Lectures & Labs ==
-
<br><u>Day 1:</u> [https://ccrma.stanford.edu/workshops/mir2010/CCRMA_MIR_2010_Lecture_1.pdf Lecture 1 Slides]
+
<br><u>Day 1:</u> [https://ccrma.stanford.edu/workshops/mir2011/CCRMA_2011_day1.pdf Lecture 1 Slides]
* '''Presenters: Jay LeBoeuf & Rebecca Fiebrink'''
* '''Presenters: Jay LeBoeuf & Rebecca Fiebrink'''
* CCRMA Introduction - (Carr/Sasha).  CCRMA Tour.
* CCRMA Introduction - (Carr/Sasha).  CCRMA Tour.
Line 46: Line 48:
* Application: Instrument recognition and drum transcription / Using simple heuristics and thresholds (i.e. "Why do we need machine learning?")  
* Application: Instrument recognition and drum transcription / Using simple heuristics and thresholds (i.e. "Why do we need machine learning?")  
-
 
+
<br><u>Lab 1:</u> <br>
-
<br><u>Lab 1:</u> [http://ccrma.stanford.edu/workshops/mir2010/Lab1_2010.pdf Lab 1 -"Manipulating audio slices"] <br>
+
* [https://ccrma.stanford.edu/workshops/mir2011/Lab_1_2011.pdf Lab 1 - Basic Feature Extraction and Classification] <br>
 +
* [http://ccrma.stanford.edu/workshops/mir2011/weka_lab1.pdf Getting started with Weka]
 +
* [https://ccrma.stanford.edu/workshops/mir2011/Wekinator_lab_2011.pdf Wekinator Lab]
 +
* Overview of Weka & the Wekinator
 +
** [http://www.cs.waikato.ac.nz/ml/weka/ Weka home]
 +
** [http://code.google.com/p/wekinator/ Wekinator on Google code] and [http://wiki.cs.princeton.edu/index.php/ChucK/Wekinator/Instructions instructions]
Students who need a personal tutorial of Matlab or audio signal processing will split off and received small group assistance to bring them up to speed.
Students who need a personal tutorial of Matlab or audio signal processing will split off and received small group assistance to bring them up to speed.
* Background for students needing a refresher:
* Background for students needing a refresher:
Line 53: Line 60:
** [http://ccrma.stanford.edu/workshops/mir2009/Lab0/lab0.html Fundamentals of Matlab]
** [http://ccrma.stanford.edu/workshops/mir2009/Lab0/lab0.html Fundamentals of Matlab]
** [http://ccrma.stanford.edu/workshops/mir2009/Lab1/lab1.html Fundamentals of Digital Audio Signal Processing (FFT, STFT, Windowing, Zero-padding, 2-D Time-frequency representation)]
** [http://ccrma.stanford.edu/workshops/mir2009/Lab1/lab1.html Fundamentals of Digital Audio Signal Processing (FFT, STFT, Windowing, Zero-padding, 2-D Time-frequency representation)]
-
Weka
 
-
* [http://ccrma.stanford.edu/workshops/mir2010/weka_lab1.pdf Getting started with Weka]
 
-
* [https://ccrma.stanford.edu/workshops/mir2010/Lab4Wekinator.pdf Wekinator Lab]
 
-
* Overview of Weka & the Wekinator
 
-
** [http://www.cs.waikato.ac.nz/ml/weka/ Weka home]
 
-
** [http://code.google.com/p/wekinator/ Wekinator on Google code] and [http://wiki.cs.princeton.edu/index.php/ChucK/Wekinator/Instructions instructions]
 
* REMINDER: Save all your work, because you may want to build on it in subsequent labs.
* REMINDER: Save all your work, because you may want to build on it in subsequent labs.
-
<br><u>Day 2:</u>  
+
<br><u>Day 2:</u> [https://ccrma.stanford.edu/workshops/mir2011/CCRMA_2011_day2.pdf Lecture 2 Slides]
 +
[http://ccrma.stanford.edu/workshops/mir2011/BeatReferences.pdf A list of beat tracking references cited]
* '''Presenters: Leigh Smith & Stephen Pope'''
* '''Presenters: Leigh Smith & Stephen Pope'''
-
* Overview: Signal Analysis and Feature Extraction for MIR Applications
+
* Overview: Signal Analysis and Feature Extraction for MIR Applications (Historical: http://quod.lib.umich.edu/cgi/p/pod/dod-idx?c=icmc;idno=bbp2372.1999.356)
* MIR Application Design
* MIR Application Design
** Audio input, analysis
** Audio input, analysis
Line 72: Line 74:
* Windowed Feature Extraction
* Windowed Feature Extraction
** I/O and analysis loops
** I/O and analysis loops
-
* Feature-vector design (http://www.create.ucsb.edu/~stp/PostScript/PopeHolmKouznetsov_icmc2.pdf)
+
* Feature-vector design (Overview: http://www.create.ucsb.edu/~stp/PostScript/PopeHolmKouznetsov_icmc2.pdf)
** Kinds/Domains of Features
** Kinds/Domains of Features
** Application Requirements (labeling, segmentation, etc.)
** Application Requirements (labeling, segmentation, etc.)
Line 86: Line 88:
* Other Feature domains
* Other Feature domains
** Wavelets, LPC
** Wavelets, LPC
-
* Onset-detection: Many Technique
+
* Onset-detection: Many Techniques
** Time-domain differences
** Time-domain differences
** Spectral-domain differences
** Spectral-domain differences
Line 97: Line 99:
** Onset-detection vs Beat-detection
** Onset-detection vs Beat-detection
*** The Onset Detection Function
*** The Onset Detection Function
-
** Beat Histograms
+
** Approaches to beat tracking & Meter estimation
-
** Fluctuation Patterns
+
-
** Meter estimation
+
-
** Approaches to beat tracking
+
-
*** Comb filters
+
*** Autocorrelation
*** Autocorrelation
*** Beat Spectrum measures
*** Beat Spectrum measures
-
*** Multiresolution (Wavelet)
+
*** Multi-resolution (Wavelet)
-
** Octave errors in beat tracking
+
** Beat Histograms
 +
** Fluctuation Patterns
 +
** Joint estimation of downbeat and chord change
* Applications
* Applications
** Song clustering based on a variety of feature vectors
** Song clustering based on a variety of feature vectors
Line 113: Line 113:
* MATLAB/Weka code for sound clustering with a flexible feature vector
* MATLAB/Weka code for sound clustering with a flexible feature vector
* C++ API examples Marsyas, Aubio, libExtract - pre-built examples to read and customize
* C++ API examples Marsyas, Aubio, libExtract - pre-built examples to read and customize
-
* Extract CAL 500 per-song features to .mat or .csv using features from today.  This will be used on lab for Friday
+
* Extract CAL 500 per-song features to .mat or .csv using features from today.  This will be used on lab for Friday. Copy it from the folder ccrma-gate.stanford.edu:/usr/ccrma/workshops/mir2011/cal500.tar (beware it's a 2Gb .tar file!) or grab the AIFF versions from ccrma-gate.stanford.edu:/usr/ccrma/workshops/mir2011/cal500_aiffs.tar (that's 16 GB)
-
 
+
* Down-loads
 +
**  [https://ccrma.stanford.edu/workshops/mir2011/MAT240F-Reader.zip UCSB MAT 240F Reader]
 +
**  [https://ccrma.stanford.edu/workshops/mir2011/MAT240F-Code.zip UCSB MAT 240F Code]
 +
**  [https://ccrma.stanford.edu/workshops/mir2011/MAT240F-Sounds.zip UCSB MAT 240F Sounds]
 +
**  [https://ccrma.stanford.edu/workshops/mir2011/ODF.zip Onset Detection Function example code in Octave/Matlab]
 +
* Notes on c-API configuration
 +
** FFTW
 +
./configure --help<br>
 +
./configure --enable-float
 +
** libSndFile
 +
./configure --disable-external-libs --disable-sqlite
 +
** CAL500 decoding
 +
for i in *.mp3; do echo $i; afconvert -d BEI16@44100 -f AIFF "$i"; done
<br><u>Day 3</u>  
<br><u>Day 3</u>  
* '''Presenters: Stephen Pope & Steve Tjoa'''
* '''Presenters: Stephen Pope & Steve Tjoa'''
 +
* [http://up.stevetjoa.com/tjoa20110629ccrma.pdf Lecture and Lab 3 Slides by Steve Tjoa]
 +
* [https://ccrma.stanford.edu/workshops/mir2011/CCRMA_2011_day3.pdf Lecture 3 Slides]
* Overview: 2nd-Stage Processing and Post-processing in MIR Applications
* Overview: 2nd-Stage Processing and Post-processing in MIR Applications
-
* 2nd-Stage Processin
+
* 2nd-Stage Processing
** Thresholds and Data Pruning
** Thresholds and Data Pruning
** Perceptual Mapping
** Perceptual Mapping
Line 155: Line 169:
<br><u>Day 4:</u>  
<br><u>Day 4:</u>  
* '''Presenters: George Tzanetakis'''
* '''Presenters: George Tzanetakis'''
-
* Features: Pitch, Chroma
+
* [http://ccrma.stanford.edu/workshops/mir2011/ccrma_2011_pitch_reps.pdf Pitch Representation Slides (.pdf)]
-
* Classification: Alignment
+
* [http://ccrma.stanford.edu/workshops/mir2011/ccrma_2011_pitch_reps.pptx Pitch Representation Slides (.pptx)]
-
* Application: Cover Song ID / Music Collections
+
* Features:  
-
 
+
** Monophonic Pitch Detection
 +
** Polyphonic Pitch Detection
 +
** Pitch representations (Tuning Histograms, Pitch and Pitch Class Profiles, Chroma)
 +
* Analysis:  
 +
** Dynamic Time Warping
 +
** Hidden Markov Models
 +
** Harmonic Analysis/Chord and Key Detection
 +
* Applications
 +
** Audio-Score Alignment  
 +
** Cover Song Detection
 +
** Analysis of religious cantillation (Computational Ethnomusicology)
 +
** Query-by-humming
 +
** Music Transcription
 +
* Tools
 +
** Marsyas
 +
** Python/NumPy/Matplotlib
<br><u>Lab 4: </u>
<br><u>Lab 4: </u>
 +
* Marsyas compilation
 +
** Instructions for CCRMA Machines [http://ccrma.stanford.edu/workshops/mir2011/marsyas_ccrma2011.pdf marsyas_ccrma2011.pdf]
 +
*** SKT: If you get an error about Python.h, install the package python2.7-dev (for version 2.7).
 +
* Marsyas tour
 +
* Plotting and prototyping using the Marsyas Python bindings
 +
* Writing some C++ Marsyas code
 +
* DTW in Matlab [http://labrosa.ee.columbia.edu/matlab/dtw/ Dan Ellis DTW Matlab example]
-
<br><u>Day 5:</u> [http://ccrma.stanford.edu/workshops/mir2010/CCRMA_MIR_2010_Lecture5.pdf]
+
 
-
* '''Presenters: Doug Eck'''
+
<br><u>Day 5:</u>  
-
* Application: Recommender
+
* '''Presenters: Douglas Eck'''
-
* Autotagging using CAL500.
+
* [http://ccrma.stanford.edu/workshops/mir2011/CCRMA_2011_day5.pdf Day 5 Slides (.pdf)]
 +
* Music Recomendation
 +
** Overview of music recommendation. What's hard about it.
 +
** Some statistics and observations about the music industry and the need for recomendation.
 +
** Point-Counterpoint: Should we bother with content-based analysis.
 +
* Autotagging
 +
** Features for autotagging (some of this will be review, given days 1 through 4.)
 +
** Demos of clustering for different types of acoustic features.
 +
** Training Data.
 +
** Classifiers (focus on AdaBoost)
 +
** Feature selection.
 +
** Evaluation with lots of examples.
 +
* Time permitting:
 +
** Advanced features
 +
** Sparse coding
 +
** Using musical structure.
<br><u>Lab 5</u>
<br><u>Lab 5</u>
 +
See [[MIR_workshop_2011_day5_lab]] for a full description.  Here is a summary:
 +
 +
* The basics (some Python code available to help).
 +
** Calculate acoustic features on CAL500 dataset (students should have already done this.)
 +
** Read in user tag annotations from same dataset provided by UCSD.
 +
** Build similarity matrix based on word vectors derived from these annotations.
 +
** Query similarity matrix with a track to get top hits based on cosine distance.
 +
** Build second similarity matrix using acoustic features.
 +
** Query this similarity matrix with track to get top hits based on cosine distance.
 +
 +
* Extra (I didn't write code for this, but can help students find examples).
 +
** Query the EchoNest for additional acoustic features and compare to yours.
 +
** Use the CAL500 user annotations as ground truth and evaluate your audio features (ROC curve or some precision measure).
 +
** Compare a 2D visualization of acoustic features versus UCSD user annotations.
 +
<br>
<br><u>Bonus Lab material</u>
<br><u>Bonus Lab material</u>
* Insert your bonus lab materials here...
* Insert your bonus lab materials here...
Line 198: Line 264:
== Supplemental papers and information for the lectures...==
== Supplemental papers and information for the lectures...==
-
[http://cm-wiki.stanford.edu/wiki/MIR_workshop_2008_notes Explanations, tutorials, code demos, recommended papers here - for each topic....]
+
* [http://cm-wiki.stanford.edu/wiki/MIR_workshop_2008_notes Explanations, tutorials, code demos, recommended papers here - for each topic....]
 +
* [http://ccrma.stanford.edu/workshops/mir2011/BeatReferences.pdf A list of beat tracking references cited]
== Past CCRMA MIR Workshops and lectures==  
== Past CCRMA MIR Workshops and lectures==  

Current revision as of 20:56, 29 July 2011

Intelligent Audio Systems: Foundations and Applications of Music Information Retrieval

Contents

Logistics

Workshop Title: "Intelligent Audio Systems: Foundations and Applications of Music Information Retrieval"

  • 9-5 PM. Mon, 06/27/2011 - Fri, 07/01/2011
  • Instructors:
- Jay LeBoeuf, Imagine Research 
- Rebecca Fiebrink, Princeton University
- Douglas Eck, Google Research Google
- Stephen Pope,  Imagine Research 
- Steve Tjoa, University of Maryland / Imagine Research 
- Leigh Smith,  Imagine Research 
- George Tzanetakis, Google Research, University of Victoria
  • Participants:
- Chris Colatos, Jeff Albert, Kamlesh Lakshminarayanan, Sean Zhang, Eli Stine, David Bird, Gina Collecchia, Dekun Zou, Bill Paseman, John Amuedo.
- See photo here: CCRMA MIR Workshop 2011 Participants

Abstract

How would you "Google for audio", provide music recommendations based your MP3 files, or have a computer "listen" and understand what you are playing? This workshop will teach the underlying ideas, approaches, technologies, and practical design of intelligent audio systems using Music Information Retrieval (MIR) algorithms.

MIR is a highly-interdisciplinary field bridging the domains of digital audio signal processing, pattern recognition, software system design, and machine learning. Simply put, MIR algorithms allow a computer to "listen" and "understand or make sense of" audio data, such as MP3s in a personal music collection, live streaming audio, or gigabytes of sound effects, in an effort to reduce the semantic gap between high-level musical information and low-level audio data. In the same way that listeners can recognize the characteristics of sound and music - tempo, key, chord progressions, genre, or song structure - MIR algorithms are capable of recognizing and extracting this information, enabling systems to perform extensive sorting, searching, music recommendation, metadata generation, transcription, and even aiding/generating real-time performance.

This workshop is intended for: students, researchers, and industry audio engineers who are unfamiliar with the field of Music Information Retrieval (MIR). We will demonstrate the myriad of exciting technologies enabled by the fusion of basic signal processing techniques with machine learning and pattern recognition. Lectures will cover topics such as low-level feature extraction, generation of higher-level features such as chord estimations, audio similarity clustering, search, and retrieval techniques, and design and evaluation of machine classification systems. The presentations will be applied, multimedia-rich, overview of the building blocks of modern MIR systems. Our goal is to make the understanding and application of highly-interdisciplinary technologies and complex algorithms approachable.

Knowledge of basic digital audio principles is required. Familiarity with Matlab is desired. Students are highly encouraged to bring their own audio source material for course labs and demonstrations.

Workshop structure: The workshop will consist of half-day lectures, half-day supervised lab sessions, demonstrations, and discussions. Labs will allow students to design basic ground-up "intelligent audio systems", leveraging existing MIR toolboxes, programming environments, and applications. Labs will include creation and evaluation of basic instrument recognition, transcription, and real-time audio analysis systems.

Lectures & Labs


Day 1: Lecture 1 Slides

  • Presenters: Jay LeBoeuf & Rebecca Fiebrink
  • CCRMA Introduction - (Carr/Sasha). CCRMA Tour.
  • Introduction to MIR (What is MIR? Why are people interested? Commercial Applications of MIR)
  • A brief history of MIR
  • Overview of a basic MIR system architecture
  • Timing and Segmentation: Frames, Onsets
  • Features: ZCR, Spectral moments; Scaling of feature data
  • Classification: Instance-based classifiers (k-NN)
  • Information Retrieval Basics
    • Classifier evaluation (Cross-validation, training and test sets)
    • IR Evaluation Metrics (precision, recall, f-measure, AROC,...)
  • Application: Instrument recognition and drum transcription / Using simple heuristics and thresholds (i.e. "Why do we need machine learning?")


Lab 1:

Students who need a personal tutorial of Matlab or audio signal processing will split off and received small group assistance to bring them up to speed.

  • REMINDER: Save all your work, because you may want to build on it in subsequent labs.


Day 2: Lecture 2 Slides A list of beat tracking references cited

  • Presenters: Leigh Smith & Stephen Pope
  • Overview: Signal Analysis and Feature Extraction for MIR Applications (Historical: http://quod.lib.umich.edu/cgi/p/pod/dod-idx?c=icmc;idno=bbp2372.1999.356)
  • MIR Application Design
    • Audio input, analysis
    • Statistical/perceptual processing
    • Data storage
    • Post-processing
  • Windowed Feature Extraction
    • I/O and analysis loops
  • Feature-vector design (Overview: http://www.create.ucsb.edu/~stp/PostScript/PopeHolmKouznetsov_icmc2.pdf)
    • Kinds/Domains of Features
    • Application Requirements (labeling, segmentation, etc.)
  • Time-domain features (MPEG-7 Audio book ref)
    • RMS, Peak, LP/HP RMS, Dynamic range, ZCR
  • Frequency-domain features
    • Spectrum, Spectral bins
    • Spectral measures (statistical moments)
    • Pitch-estimation and tracking
    • MFCCs
  • Spatial-domain features
    • M/S Encoding, Surround-sound Processing Frequency-dependent spatial separation, LCR sources
  • Other Feature domains
    • Wavelets, LPC
  • Onset-detection: Many Techniques
    • Time-domain differences
    • Spectral-domain differences
    • Perceptual data-warping
    • Adaptive onset detection
  • Beat-finding and Tempo Derivation
    • IOIs and Beat Regularity, Rubato
      • Tatum, Tactus and Meter levels
      • Tempo estimation
    • Onset-detection vs Beat-detection
      • The Onset Detection Function
    • Approaches to beat tracking & Meter estimation
      • Autocorrelation
      • Beat Spectrum measures
      • Multi-resolution (Wavelet)
    • Beat Histograms
    • Fluctuation Patterns
    • Joint estimation of downbeat and chord change
  • Applications
    • Song clustering based on a variety of feature vectors
    • PCA of feature spaces using Weka


Lab 2:

  • Feature extraction and flexible feature vectors in MATLAB, Marsyas, Aubio, libExtract
  • MATLAB/Weka code for sound clustering with a flexible feature vector
  • C++ API examples Marsyas, Aubio, libExtract - pre-built examples to read and customize
  • Extract CAL 500 per-song features to .mat or .csv using features from today. This will be used on lab for Friday. Copy it from the folder ccrma-gate.stanford.edu:/usr/ccrma/workshops/mir2011/cal500.tar (beware it's a 2Gb .tar file!) or grab the AIFF versions from ccrma-gate.stanford.edu:/usr/ccrma/workshops/mir2011/cal500_aiffs.tar (that's 16 GB)
  • Down-loads
  • Notes on c-API configuration
    • FFTW

./configure --help
./configure --enable-float

    • libSndFile

./configure --disable-external-libs --disable-sqlite

    • CAL500 decoding

for i in *.mp3; do echo $i; afconvert -d BEI16@44100 -f AIFF "$i"; done
Day 3

  • Presenters: Stephen Pope & Steve Tjoa
  • Lecture and Lab 3 Slides by Steve Tjoa
  • Lecture 3 Slides
  • Overview: 2nd-Stage Processing and Post-processing in MIR Applications
  • 2nd-Stage Processing
    • Thresholds and Data Pruning
    • Perceptual Mapping
    • Data Reduction: Averaging, GMMs, Running Averages
    • Feature-data-smoothing: de-spiking, sticky values, filter, etc.
  • Segmentation of music and non-musical audio
    • Segmentation based on islands of similar features
    • Segmentation based on regular difference peaks
    • Segmentation based on labeling
  • Post-processing: What are we doing?
    • Storing Feature Data: SQL, JSON, XML, etc.
    • Classification/Clustering/Transcription/Labeling
  • Classification: KNN vs SVM training and testing
    • SVM tools and APIs
  • Clustering vs Classification: Tree-based systems
  • Audio Transcription: Onsets and per-onset features
  • Other applications: source separation, similarity match, search, etc.
  • Classification/estimation in the presence of polyphony
    • Try basic approach on a musical mixture.
    • How well does it perform?
    • What do we do to improve its performance? ICA, NMF, K-SVD.
    • Matrix representations of data: spectrogram, chromagram, timbregram, etc.
    • Methods to improve NMF/K-SVD under heavy harmonic overlap
  • Applications
    • Feature vector pruning
    • Segmentation examples
    • SVMs for classification
    • Multipitch estimation, source separation, denoising


Lab 3:

  • 2nd-Stage Processing
  • SVM tools
  • Classification examples

  • If you finish early, see the "bonus labs" section below.

Day 4:
  • Presenters: George Tzanetakis
  • Pitch Representation Slides (.pdf)
  • Pitch Representation Slides (.pptx)
  • Features:
    • Monophonic Pitch Detection
    • Polyphonic Pitch Detection
    • Pitch representations (Tuning Histograms, Pitch and Pitch Class Profiles, Chroma)
  • Analysis:
    • Dynamic Time Warping
    • Hidden Markov Models
    • Harmonic Analysis/Chord and Key Detection
  • Applications
    • Audio-Score Alignment
    • Cover Song Detection
    • Analysis of religious cantillation (Computational Ethnomusicology)
    • Query-by-humming
    • Music Transcription
  • Tools
    • Marsyas
    • Python/NumPy/Matplotlib

Lab 4:
  • Marsyas compilation
    • Instructions for CCRMA Machines marsyas_ccrma2011.pdf
      • SKT: If you get an error about Python.h, install the package python2.7-dev (for version 2.7).
  • Marsyas tour
  • Plotting and prototyping using the Marsyas Python bindings
  • Writing some C++ Marsyas code
  • DTW in Matlab Dan Ellis DTW Matlab example

Day 5:
  • Presenters: Douglas Eck
  • Day 5 Slides (.pdf)
  • Music Recomendation
    • Overview of music recommendation. What's hard about it.
    • Some statistics and observations about the music industry and the need for recomendation.
    • Point-Counterpoint: Should we bother with content-based analysis.
  • Autotagging
    • Features for autotagging (some of this will be review, given days 1 through 4.)
    • Demos of clustering for different types of acoustic features.
    • Training Data.
    • Classifiers (focus on AdaBoost)
    • Feature selection.
    • Evaluation with lots of examples.
  • Time permitting:
    • Advanced features
    • Sparse coding
    • Using musical structure.

Lab 5 See MIR_workshop_2011_day5_lab for a full description. Here is a summary:
  • The basics (some Python code available to help).
    • Calculate acoustic features on CAL500 dataset (students should have already done this.)
    • Read in user tag annotations from same dataset provided by UCSD.
    • Build similarity matrix based on word vectors derived from these annotations.
    • Query similarity matrix with a track to get top hits based on cosine distance.
    • Build second similarity matrix using acoustic features.
    • Query this similarity matrix with track to get top hits based on cosine distance.
  • Extra (I didn't write code for this, but can help students find examples).
    • Query the EchoNest for additional acoustic features and compare to yours.
    • Use the CAL500 user annotations as ground truth and evaluate your audio features (ROC curve or some precision measure).
    • Compare a 2D visualization of acoustic features versus UCSD user annotations.


Bonus Lab material

software, libraries, examples

Applications & Environments

Machine Learning Libraries & Toolboxes

Optional Toolboxes

Supplemental papers and information for the lectures...

Past CCRMA MIR Workshops and lectures

References for additional info

Recommended books:

  • Data Mining: Practical Machine Learning Tools and Techniques, Second Edition by Ian H. Witten , Eibe Frank (includes software)
  • Netlab by Ian T. Nabney (includes software)
  • Signal Processing Methods for Music Transcription, Klapuri, A. and Davy, M. (Editors)
  • Computational Auditory Scene Analysis: Principles, Algorithms, and Applications, DeLiang Wang (Editor), Guy J. Brown (Editor)
  • Speech and Audio Signal Processing:Processing and perception of speech and music Ben Gold & Nelson Morgan, Wiley 2000

Prerequisite / background material:

Papers:

Other books:

  • Pattern Recognition and Machine Learning (Information Science and Statistics) by Christopher M. Bishop
  • Neural Networks for Pattern Recognition, Christopher M. Bishop, Oxford University Press, 1995.
  • Pattern Classification, 2nd edition, R Duda, P Hart and D Stork, Wiley Interscience, 2001.
  • "Artificial Intelligence: A Modern Approach" Second Edition, Russell R & Norvig P, Prentice Hall, 2003.
  • Machine Learning, Tom Mitchell, McGraw Hill, 1997.

Interesting Links:

Audio Source Material

OLPC Sound Sample Archive (8.5 GB) [1]

http://www.tsi.telecom-paristech.fr/aao/en/category/database/

RWC Music Database (n DVDs) [available in Stanford Music library]

RWC - Sound Instruments Table of Contents

http://staff.aist.go.jp/m.goto/RWC-MDB/rwc-mdb-i.html

Univ or Iowa Music Instrument Samples

https://ccrma.stanford.edu/wiki/MIR_workshop_2008_notes#Research_Databases_.2F_Collections_of_Ground_truth_data_and_copyright-cleared_music

MATLAB Utility Scripts

http://ccrma.stanford.edu/~kglee/kaist_summer2008_special_lecture/

Personal tools