• Corpus ID: 1352334


  author={Giovanni Costantini and Fabio Massimo Frattale Mascioli and Patrizio Antici},
The classification of single musical sources is an essential step in order to obtain the source separation and the automatic transcription of polyphonic music. In this paper, we present a first experience of recognition of five different musical instruments (clarinet, flute, oboe, saxophone and violin). For such task, a nonexclusive classifier capable of fuzzy decisions is especially suitable, due to the inevitable overlaps among data. We used two different neuro-fuzzy classifier for… 

Figures and Tables from this paper

Recognition of musical instruments by statistical classification

A hierarchical scale-based approach is proposed for the classification of single musical sources based on FFT and QFT (Q-constant Frequency Transform) for feature extraction and data set preparation, carrying out good results.

Adaptive resolution min-max classifiers

Two new learning algorithms for fuzzy min-max neural classifiers are proposed: the adaptive resolution classifier (ARC) and its pruning version (PARC), which allow to achieve networks with a remarkable generalization capability.

A recursive algorithm for fuzzy min-max networks

  • A. RizziM. PanellaF. MascioliG. Martinelli
  • Computer Science
    Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks. IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium
  • 2000
An algorithm to train min-max neural models is proposed. It is based on the adaptive resolution classifier (ARC) technique, which overcomes some undesired properties of the original Simpson's (1992)



A Blackboard System for Automatic Transcription of Simple Polyphonic Music

A novel computational system has been constructed which is capable of transcribing piano performances of four-voice Bach chorales written in the style of 18th century counterpoint. The system is

Algorithm for extraction of pitch and pitch salience from complex tonal signals

A procedure is described for the automatic extraction of the various pitch percepts which may be simultaneously evoked by complex tonal stimuli. The procedure is based on the theory of virtual pitch,

A nonexclusive classification system based on co-operative fuzzy clustering

This paper presents an algorithm that solves a nonexclusive k-class problem by the co-operation of k independent clustering systems, and proposes a fuzzy classification quality (FCQ) measure.

A Review of Probabilistic, Fuzzy, and Neural Models for Pattern Recognition

  • J. Bezdek
  • Computer Science
    J. Intell. Fuzzy Syst.
  • 1993
The basic ideas of and some synergisms between probabilistic, fuzzy, and computational neural networks models as they apply to pattern recognition are discussed.

Fuzzy Min-Max Neural Networks-Part 1 : Classification

A neural network classifier that creates classes by aggregating several smaller fuzzy sets into a single fuzzy set class that can add new pattern classes on the fly, refine existing pattern classes as new information is received, and it uses simple operations that allow for quick execution is described.

Neuro-fuzzy modeling and control

The essential part of neuro-fuzzy synergisms comes from a common framework called adaptive networks, which unifies both neural networks and fuzzy models, which possess certain advantages over neural networks.

Physical Correlates of Brass‐Instrument Tones

Time‐dependent Fourier analyses of the attack transients and steady states of about 900 trumpet, trombone, tuba, and French‐horn tones yielded the amplitudes and relative phases of the first 11

Fuzzy min-max neural networks - Part 2: Clustering

This paper will provide some background concerning the development of the fuzzy min-max clustering neural network and provide a comparison with similar work that has recently emerged and a brief description of fuzzy sets, pattern clustering, and their synergistic combination is presented.

A geometric approach to cluster validity for normal mixtures

Viewing mixture decomposition as probabilistic clustering as opposed to parametric estimation enables both fuzzy and crisp measures of cluster validity for this problem, and uses the expectation-maximization algorithm to find clusters in the data.