Please note:

This page has not been properly updated since 2001. A more recent overview of our work can be found here

	RESEARCH TOPICS	RESEARCH ITEMS INCLUDED
more transcription oriented more classification oriented	MULTI-PITCH ESTIMATION Anssi Klapuri Estimation of the pitch of multiple simultaneous sounds Docs & demos	PITCH MODELS Anssi Klapuri Review and development of algorithms and models that can be used to estimate pitch	SPECTRUM SEPARATION Anssi Klapuri Separation of the spectra of simultaneous sounds	MUSICAL POLYPHONIES Anssi Klapuri Resolving a mixture of several harmonic sounds that may be harmonically related.
	RHYTHM PROCESSING Jarno Seppänen Jukka Sillanpää Estimation of the rhythmic features in musical signals	QUANTUM, BEAT, QUANTIZATION Jarno Seppänen Finding the metrical grid and foot tapping rate in a musical performance. Docs & demos	DRUM TYPE RECOGNITION Jukka Sillanpää Drum instrument detection and recognition Docs & demos	ONSET DETECTION Anssi Klapuri Detection of the beginnings of dicrete acoustic events in a signal.
	INSTRUMENT RECOGNITION Antti Eronen Pitch-independent recognition of a musical instrument from an acoustic signal Docs & demos	TIMBRE FEATURES Antti Eronen Literature review, evaluation and testing of features for timbre classification	GENERIC CLASSIFICATION Classification of a new instance, given training examples
	SPECTRUM MODELING Tuomas Virtanen Modeling sounds with sinusoids plus noise. In between content analysis and coding. Docs & demos	SINUSOID TRACKING Tuomas Virtanen Finding continuous sinusoids in a signal.	RESIDUAL MODELING Tuomas Virtanen Subtraction of known sinusoids from a signal and modeling the residual noise.
	MUSIC CLASSIFICATION Jarno Seppänen Jukka Sillanpää Classification of musical signals into subclasses, according to a musical style or other criteria	RHYTHM PROCESSING Jarno Seppänen Jukka Sillanpää See above.	SIGNAL FEATURES Arnaud Philibert See below.	INSTRUMENT RECOGNITION Antti Eronen See above.
	SPEECH / MUSIC DISCRIMINATION Arnaud Philibert Classification of a given acoustic signal as speech or music. Docs & demos	SIGNAL FEATURES Arnaud Philibert Literature review, implementation, and simulations of feature for speech/ music discrimination	GENERIC CLASSIFICATION See above.

Contact person for the music content analysis team is Anssi Klapuri (, klap @ cs tut fi).