ACE: A Framework for optimizing music classification Cory McKay Rebecca Fiebrink Daniel McEnnis Beinan Li Ichiro Fujinaga Music Technology Area Faculty.

Slides:



Advertisements
Similar presentations
Florida International University COP 4770 Introduction of Weka.
Advertisements

Idaho National Engineering and Environmental Laboratory What is a Framework? Web Service? Why do you need them? Wayne Simpson November.
Automatic Music Classification Cory McKay. 2/47 Introduction Many areas of research in music information retrieval (MIR) involve using computers to classify.
Copyright © 2014 Pearson Education, Inc. Publishing as Prentice Hall
File Management Chapter 3
Weka & Rapid Miner Tutorial By Chibuike Muoh. WEKA:: Introduction A collection of open source ML algorithms – pre-processing – classifiers – clustering.
ClearTK: A Framework for Statistical Biomedical Natural Language Processing Philip Ogren Philipp Wetzler Department of Computer Science University of Colorado.
Machine Learning Group University College Dublin 4.30 Machine Learning Pádraig Cunningham.
Department of Computer Science, University of Waikato, New Zealand Eibe Frank WEKA: A Machine Learning Toolkit The Explorer Classification and Regression.
Learning Programs Danielle and Joseph Bennett (and Lorelei) 4 December 2007.
Slide 1 of 9 Presenting 24x7 Scheduler The art of computer automation Press PageDown key or click to advance.
 The Weka The Weka is an well known bird of New Zealand..  W(aikato) E(nvironment) for K(nowlegde) A(nalysis)  Developed by the University of Waikato.
JSymbolic and ELVIS Cory McKay Marianopolis College Montreal, Canada.
Attention Deficit Hyperactivity Disorder (ADHD) Student Classification Using Genetic Algorithm and Artificial Neural Network S. Yenaeng 1, S. Saelee 2.
Introduction to the Enterprise Library. Sounds familiar? Writing a component to encapsulate data access Building a component that allows you to log errors.
CSCI 347 – Data Mining Lecture 01 – Course Overview.
Data Exchange Tools (DExT) DExT PROJECTAN OPEN EXCHANGE FORMAT FOR DATA enables long-term preservation and re-use of metadata,
SCRAM Software Configuration, Release And Management Background SCRAM has been developed to enable large, geographically dispersed and autonomous groups.
ENN: Extended Nearest Neighbor Method for Pattern Recognition
Issues with Data Mining
Comparing the Parallel Automatic Composition of Inductive Applications with Stacking Methods Hidenao Abe & Takahira Yamaguchi Shizuoka University, JAPAN.
Chapter 7 Web Content Mining Xxxxxx. Introduction Web-content mining techniques are used to discover useful information from content on the web – textual.
Chapter 6 : Software Metrics
Appendix: The WEKA Data Mining Software
1 Research Groups : KEEL: A Software Tool to Assess Evolutionary Algorithms for Data Mining Problems SCI 2 SMetrology and Models Intelligent.
JSymbolic Cedar Wingate MUMT 621 Professor Ichiro Fujinaga 22 October 2009.
Presented by Abirami Poonkundran.  Introduction  Current Work  Current Tools  Solution  Tesseract  Tesseract Usage Scenarios  Information Flow.
Versus JEDEC STAPL Comparison Toolkit Frank Toth February 20, 2000.
Use of Hierarchical Keywords for Easy Data Management on HUBzero HUBbub Conference 2013 September 6 th, 2013 Gaurav Nanda, Jonathan Tan, Peter Auyeung,

A Language Independent Method for Question Classification COLING 2004.
Project 1: Machine Learning Using Neural Networks Ver 1.1.
Machine Learning with Weka Cornelia Caragea Thanks to Eibe Frank for some of the slides.
Today Ensemble Methods. Recap of the course. Classifier Fusion
Ensemble Methods: Bagging and Boosting
Automatic music classification and the importance of instrument identification Cory McKay and Ichiro Fujinaga Music Technology Area Faculty of Music McGill.
W E K A Waikato Environment for Knowledge Analysis Branko Kavšek MPŠ Jožef StefanNovember 2005.
Explorations into Internet Distributed Computing Kunal Agrawal, Ang Huey Ting, Li Guoliang, and Kevin Chu.
Object-Oriented Software Engineering using Java, Patterns &UML. Presented by: E.S. Mbokane Department of System Development Faculty of ICT Tshwane University.
interactive logbook Paul Kiddie, Mike Sharples et al. The Development of an Application to Enhance.
Intelligent Database Systems Lab N.Y.U.S.T. I. M. Externally growing self-organizing maps and its application to database visualization and exploration.
BOĞAZİÇİ UNIVERSITY DEPARTMENT OF MANAGEMENT INFORMATION SYSTEMS MATLAB AS A DATA MINING ENVIRONMENT.
McGill University > Schulich School of Music > Music Technology > MUMT 611 j j MusicMetaManager j j Cory McKay Jason A. Hockman part of the jMIR software.
Weka – A Machine Learning Toolkit October 2, 2008 Keum-Sung Hwang.
W E K A Waikato Environment for Knowledge Aquisition.
Issues in Automatic Musical Genre Classification Cory McKay.
Exploiting Named Entity Taggers in a Second Language Thamar Solorio Computer Science Department National Institute of Astrophysics, Optics and Electronics.
Digitization of the Lester S. Levy Collection of Sheet Music Ichiro Fujinaga McGill University with Michael Droettboom, Karl MacMillan, G. Sayeed Choudhury,
High Throughput and Programmable Online Traffic Classifier on FPGA Author: Da Tong, Lu Sun, Kiran Kumar Matam, Viktor Prasanna Publisher: FPGA 2013 Presenter:
1 / 22 jSymbolic Jordan Smith – MUMT 611 – 6 March 2008.
Wrapper Learning: Cohen et al 2002; Kushmeric 2000; Kushmeric & Frietag 2000 William Cohen 1/26/03.
Software Design and Architecture
1 Munther Abualkibash University of Bridgeport, CT.
Your Interactive Guide to the Digital World Discovering Computers 2012 Chapter 13 Computer Programs and Programming Languages.
Combining Models Foundations of Algorithms and Machine Learning (CS60020), IIT KGP, 2017: Indrajit Bhattacharya.
Big data classification using neural network
Recent Trends in Text Mining
LOCO Extract – Transform - Load
MultiRefactor: Automated Refactoring To Improve Software Quality
Source: Procedia Computer Science(2015)70:
Waikato Environment for Knowledge Analysis
WEKA.
Machine Learning with Weka
Project 1: Text Classification by Neural Networks
Optical Music Recognition
iSRD Spam Review Detection with Imbalanced Data Distributions
Course: Module: Lesson # & Name Instructional Material 1 of 32 Lesson Delivery Mode: Lesson Duration: Document Name: 1. Professional Diploma in ERP Systems.
Machine Learning with Weka
Lecture 10 – Introduction to Weka
What's New in eCognition 9
Presentation transcript:

ACE: A Framework for optimizing music classification Cory McKay Rebecca Fiebrink Daniel McEnnis Beinan Li Ichiro Fujinaga Music Technology Area Faculty of Music McGill University

2/25 Goals  Highlight limitations of existing pattern recognition software when applied to MIR  Present solutions to these limitations  Stress importance of standardized classification and feature extraction software  Ease of use, portability and extensibility  Present the ACE software framework  Uses meta-learning  Uses classification ensembles

3/25 Existing music classification systems  Systems often implemented with specific tasks in mind  Not extensible to general tasks  Often difficult to use for those not involved in project  Need standardized systems for a variety of MIR problems  No need to reimplement existing algorithms  More reliable code  More usable software  Facilitates comparison of methodologies  Important foundations  Marsyas (Tzanetakis & Cook 1999)  M2K (Downie 2004)

4/25 Existing general classification systems  Available general-purpose systems:  PRTools (van der Heijden et al )  Weka (Witten & Frank 2005)  Other meta-learning systems:  AST (Lindner and Studer 1999)  Metal (

5/25 Problems with existing systems  Distribution problems  Proprietary software  Not open source  Limited licence  Music-specific systems are often limited  None use meta-learning  Classifier ensembles rarely used  Interfaces not oriented towards end users  General-purpose systems not designed to meet the particular needs of music

6/25 Special needs of music classification (1)  Assign multiple classes to individual recordings  A recording may belong to multiple genres, for example  Allow classification of sub-sections and of overall recordings  Audio features often windowed  Useful for segmentation problems  Maintain logical grouping of multi-dimensional features  Musical features often consist of vectors (e.g. MFCC’s)  This relatedness can provide classification opportunities

7/25 Special needs of music classification (2)  Maintain identifying meta-data about instances  Title, performer, composer, date, etc.  Take advantage of hierarchically structured taxonomies  Humans often organize music hierarchically  Can provide classification opportunities  Interface for any user

8/25 Standardized file formats  Existing formats such as Weka’s ARFF format cannot represent needed information  Important to enable classification systems to communicate with arbitrary feature extractors  Four XML file formats that meet the above needs are described in proceedings

9/25 The ACE framework  ACE (Autonomous Classification Engine) is a classification framework that can be applied to arbitrary types of music classification  Meets all requirements presented above  Java implementation makes ACE portable and easy to install

10/25 ACE and meta-learning  Many classification methodologies available  Each have different strengths and weaknesses  Uses meta-learning to experiment with a variety of approaches  Finds approaches well suited to each problem  Makes powerful pattern recognition tools available to non- experts  Useful for benchmarking new classifiers and features

11/25 ACE Feature Extraction System Classification Methodology n Dimensionality Reduction Classification Methodology 1 Dimensionality Reduction … Model Classifications Music Recordings TaxonomyFeature Settings Extracted Features Experiment Coordinator Classifier Evaluator Trained ClassifiersStatistical Comparison of Classification Methodologies

12/25 Algorithms used by ACE  Uses Weka class libraries  Makes it easy to add or develop new algorithms  Candidate classifiers  Induction trees, naive Bayes, k-nearest neighbour, neural networks, support vector machines  Classifier parameters are also varied automatically  Dimensionality reduction  Feature selection using genetic algorithms, principal component analysis, exhaustive searches  Classifier ensembles  Bagging, boosting

13/25 Classifier ensembles  Multiple classifiers operating together to arrive at final classifications  e.g. AdaBoost (Freund and Shapire 1996)  Success rates in many MIR areas are behaving asymptotically (Aucouturier and Pachet 2004)  Classifier ensembles could provide some improvement

14/25 Musical evaluation experiments  Achieved a 95.6% success with a five-class beatbox recognition experiment (Sinyor et al. 2005)  Repeated Tindale’s percussion recognition experiment (2004)  ACE achieved 96.3% success, as compared to Tindale’s best rate of 94.9%  A reduction in error rate of 27.5%

15/25 General evaluation experiments  Applied ACE to six commonly used UCI datasets  Compared results to recently published algorithm (Kotsiantis and Pintelas 2004)

16/25 Results of UCI experiments (1) Data Set ACE's Selected Classifier Kotsiantis' Success Rate ACE's Success Rate autosAdaBoost81.70%86.30% diabetesNaïve Bayes76.60%78.00% ionosphereAdaBoost90.70%94.30% irisFF Neural Net95.60%97.30% labork-NN93.40%93.00% voteDecision Tree96.20%96.30%

17/25 Results of UCI experiments (2)  ACE performed very well  Statistical uncertainty makes it difficult to say that ACE’s results are inherently superior  ACE can perform at least as well as a state of the art algorithm with no tweaking  ACE achieved these results using only one minute per learning scheme for training and testing

18/25 Results of UCI experiments (3)  Different classifiers performed better on different datasets  Supports ACE’s experimental meta-learning approach  Effectiveness of AdaBoost (chosen 2 times out of 6) demonstrates strength of classifier ensembles

19/25 Feature extraction  ACE not tied to any particular feature extraction system  Reads Weka ARFF as well as ACE XML files  Does include two powerful and extensible feature extractors are bundled with ACE  Write Weka ARFF as well as ACE XML

20/25 jAudio  Reads: .mp3 .wav .aiff .au .snd

21/25 jSymbolic  Reads MIDI  Uses 111 Bodhidharma features

22/25 ACE’s interface  Graphical interface  Includes an on-line manual  Command-line interface  Batch processing  External calls  Java API  Open source  Well documented  Easy to extend

23/25 Current status of ACE  In alpha release  Full release scheduled for January 2006  Finalization of GUI  User constraints on training, classification and meta- learning times  Feature weighting  Expansion of candidate algorithms  Long-term  Distributed processing, unsupervised learning, blackboard systems, automatic cross-project optimization

24/25 Conclusions  Need standardized classification software able to deal with the special needs of music  Techniques such as meta-learning and classifier ensembles can lead to improved performance  ACE designed to address these issues

 Web site:  coltrane.music.mcgill.ca/ACE  