Springer, 2005. — 433 p.The emerging concept of human-centered computing or anthropomorphic approach represents a significant move towards intelligent systems, and affords a new perspective on information technology. Relationships between the human brain, mind and perception that have the potential of enhancing peoples’ cognitive performance can be found in many domains, examples of which will be shown in relation to music processing and classification. On the other hand, it would be advisable to design systems capable of imitating perceptual processes that are best adapted to specific technological problems. The objective of this monograph is to provide novel insights into perceptual mechanisms underlying the processing of sound and music in different environments. A solid understanding of these mechanisms is vital for numerous technological applications such as, for example, information retrieval from distributed musical databases. In order to investigate the cognitive mechanisms underlying music perception some soft computing methods will be used. The system proposed by the Author, based on the rough set method and fuzzy logic, provides knowledge on how humans internally represent such notions as quality and timbre and therefore it allows the human-like automatic processing of musical data. In addition, the automatically extracted knowledge on the above processes can be compared to fundamentals of hearing psychophysiology and to principles of music perception. Also other applications of hybrid decision systems to problem solving in music and acoustics will be exemplified and discussed in this book based not only on the review of some literature sources, but also on the experimental results obtained in the Multimedia System Department, Gdansk University of Technology. The aim of this book is to show examples of the implementation of computational intelligence methods in musical signal and music analysis, as well as in the classification tasks. A part of this book contains a short review of perceptual bases of hearing and music. Then methods and techniques that can be classified as computational intelligence or machine-learning are shortly introduced. The presented methods are applied in the areas considered to be most relevant to music information retrieval (MIR) and acoustics. Accordingly, methods based on such learning algorithms as neural networks, rough sets, fuzzy-logic, and genetic algorithms were conceived, implemented and tested on musical data. In addition, the abovementioned methods were applied to the analysis of musical duets, musical phrases and audio signals. Another problem discussed within the framework of this book is the ‘computing with words’ concept applied to both acoustics and psychophysiology. Perception-based analysis applied to psychophysiology focuses on the evaluation of hearing impairments. Application of neural networks to the processing of beamformer signals is another issue reviewed in this book. The last application described is devoted to the problem of audio-visual correlation search. This is based on a hybrid system consisting of rough-fuzzy and evolutionary computation methods.Introduction. Perceptual Bases of Hearing and Music Perception. Intelligent Musical Instrument Sound Classification. Cognitive Approach to Musical Data Analysis. Cognitive Processing in Acoustics. Synesthetic Analysis of Audio-Visual Data.
Чтобы скачать этот файл зарегистрируйтесь и/или войдите на сайт используя форму сверху.