RuSSIR 2013 QBSH and AFP as Two Successful Paradigms of Music Information Retrieval Jyh-Shing Roger Jang ( 張智星 ) MIR Lab, CSIE Dept National Taiwan Univ., Taiwan
Course Outline zQBSH & AFP yQuery by singing/humming yAudio Fingerprinting (aka query by exact example) zComparison yServe the same purpose yAll the others differ a lot xDifferent features xDifferent retrieval methods xDifferent bottlenecks zOur foci yFeature extraction & Comparison methods yOptimization strategies yMATLAB prototypes & system demos zCourse syllabusCourse syllabus
Course Schedule zQBSH ySept 16, 2003 xIntro. to audio signalsIntro. to audio signals xBasic audio featuresBasic audio features ySept 17, 2003 xPitch trackingPitch tracking ySept 18, 2003 xComparison methodsComparison methods xQBSH demosQBSH demos xQBSH prototype in MATLAB zAFP ySept 19, 2003 xAFP feature extraction and comparisonAFP feature extraction and comparison ySept 20, 2003 xAFP demos xAFP prototype in MATLAB
Related Resources zOnline tutorials yAudio signal processing and recognition: oks/audioSignalProcessi ng oks/audioSignalProcessi ng yData clustering and pattern recognition: oks/dcpr oks/dcpr zMATLAB toolboxes yMachine learning toolbox: tlab/toolbox/machineLea rning tlab/toolbox/machineLea rning ySAP (speech and audio processing) toolbox: tlab/toolbox/sap tlab/toolbox/sap