Carver, Lesser. (1992). "Blackboard Systems for Knowledge-Based Signal Understanding". Oppenheim and Nawab (eds.). "Symbolic and Knowledge-Based Signal Processing". Prentice Hall.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....that creates and modifies the sound elements in response to the comparison between predictions gen erated by the existing representation and the features from the front end. In the implementation, this was accomplished by a blackboard system, inspired by and indeed based upon the IPUS system [16,17]. Blackboard systems are well suited to problems of constructing abductive inferences for the causes of observed results: they support competing hierarchies of explanatory hypotheses, and, in situations where the ideal algorithm is unknown or unpredictable, can find a processing sequence based ....

N. Carver, V. Lessen "Blackboard systems for knowledge -based signal understanding," in Symbolic and Knowledge-based Signal Processing, ed. A. Oppenheim and S. Nawab, Prentice Hall, 1992.

....of data driven sound organization (upper panel) such as [Brown 1992] contrasted with the predictiondriven approach (lower panel) described in this paper. Ellis Using knowledge to organize sound 1998Jan20 5 of the system. It is implemented as a rule base (within a blackboard framework [Carver Lesser 1992]) which classifies the different kinds of prediction inadequacies depending on the elements involved. The basic operation is to adjust the parameters of existing elements, provided the deviations are not too great; when the discrepancies are larger it will add new elements or terminate existing ....

N. Carver & V. Lesser (1992), "Blackboard systems for knowledge-based signal understanding," in: A. Oppenheim & S. Nawab, eds., Symbolic and knowledge-based signal processing (Prentice Hall, NewYork).

....of the necessary knowledge, and able to act independently to solve the larger explanation problem. Knowledge sources typically co operate through a common data structure, called a blackboard. Several systems for computational auditory scene analysis have been built around blackboard architectures (Carver and Lesser, 1992; Nawab and Lesser, 1992; Cooke et al. 1993; Nakatani et al. 1998; Ellis, 1996; Klassner, 1996; Godsmark and Brown, 1999) Blackboards support an arbitrary combination of data driven and hypothesis driven activity, making them suitable for incorporating higher level knowledge of use in the ....

Carver, N. and Lesser, V. (1992), Blackboard systems for knowledge-based signal understanding, in: Symbolic and knowledge-based signal processing (eds: A.V. Oppenheim and S.H. Nawab), Prentice Hall.

....Lab Perceptual Computing Technical Report #224 (rev. 2) Submitted to the 12th International Conference on Pattern Recognition, Jerusalem, October Ellis Computer Psychoacoustic Grouping 1994jan28 2 both in outline and to some extent in detail, to the soundunderstanding blackboard systems of [Carv92] and [Nawa92] an axis of abstractionwehope to investigate and develop in later work. 1.1 The motivation for modeling Although it would certainly be useful to endow machines with the abilities to organize and interpret sound that are possessed by people, our primary motivation has been to ....

N Carver, V Lesser (1992) "Blackboard systems for knowledge-based signal understanding, " in Symbolic and knowledge-based signal processing, ed. AV Oppenheim & SH Nawab, Prentice Hall

....into a parametric classifier, since everything the classifier knows about variability in the measurements is represented in the (algorithm specific) classifier weights. It is possible, however, to combine parametric classifiers and error models under the supervision of a higher level program. Carver and Lesser (1992), for example, have developed a non speech sound recognizer using an expert system. In their system, signal processing knowledge sources contribute knowledge about the error inherent in different signal representations, and the expert system uses this knowledge to decide which representations to ....

....of discriminant error from measurement error in section 5.1 shows one way in which aggregate error models, of the sort developed in chapter 3, might be used in a larger knowledge based speech recognizer. A knowledge based recognizer built, for example, using a blackboard expert system (e.g. Carver and Lesser, 1992) usually depends on quantitative estimates of the reliability of various competing recognition hypotheses. Section 5.1 demonstrates the use of quantitative error models to predict the reliability of several binary classifiers, and of the round robin classifier built from them; similar prediction ....

Carver, N. and Lesser, V. (1992). Blackboard systems for knowledge-based signal understanding. In Oppenheim, A. V. and Nawab, S. H., editors, Symbolic and Knowledge-Based Signal Processing, pages 205--250. Prentice-Hall, Englewood Cliffs, NJ.

....problems all depend on a technology which builds, manipulates, ranks, prunes, and displays grouping hypotheses. Our efforts in this direction are partly inspired by other research on sensor interpretation problems where domain independent methods have been utilized. Particular examples include [3], 11] and [9] A. Machine Rhythm The purpose of the Machine Rhythm program is to use information about the onset, offset, pitch and loudness of each note in a musical performance to determine the rhythm of the performance. Rhythm, here, refers to the structure that is perceived by human ....

N. Carver and V. Lesser, "Blackboard systems for knowledge-based signal understanding," in Symbolic and Knowledge-Based Signal Processing, eds. A. Oppenheim and S. Nawab, New York: Prentice Hall, 1992.

....engine that creates and modifies the sound elements in response to the comparison between predictions generated by the existing representation and the features from the front end. In the implementation, this was accomplished by a blackboard system, inspired by and indeed based upon the IPUS system [16,17]. Blackboard systems are well suited to problems of constructing abductive inferences for the causes of observed results: they support competing hierarchies of explanatory hypotheses, and, in situations where the ideal algorithm is unknown or unpredictable, can find a processing sequence based ....

N. Carver, V. Lesser. "Blackboard systems for knowledge -based signal understanding," in Symbolic and Knowledge-based Signal Processing, ed. A. Oppenheim and S. Nawab, Prentice Hall, 1992.

No context found.

Norman Carver and Victor Lesser, "Blackboard Systems for Knowledge-Based Signal Understanding," in Symbolic and Knowledge-Based Signal Processing, Alan Oppenheim and Hamid Nawab, editors, Prentice Hall, 205--250, 1992.

....solving architecture as it was developed in Hearsay II. Because our principal focus is control, the other components of the blackboard architecture will be discussed only to the extent necessary to understand control issues. More complete discussions of the blackboard model can be found in [Carver92, Engelmore88, Erman80, Nii86a, Nii86b] The first subsection of this section introduces the concepts that underlie the blackboard model, while the following subsection examines the agenda based control mechanism used in HSII. 2.1 The Blackboard Model of Problem Solving In the basic model that ....

Norman Carver and Victor Lesser, "Blackboard Systems for Knowledge-Based Signal Understanding," in Symbolic and Knowledge-Based Signal Processing, Alan Oppenheim and Hamid Nawab, editors, Prentice Hall, 205--250, 1992.

No context found.

Norman Carver and Victor Lesser, "Blackboard Systems for Knowledge-Based Signal Understanding," in Symbolic and Knowledge-Based Signal Processing, Alan Oppenheim and Hamid Nawab, editors, Prentice Hall, 205--250, 1992.

No context found.

Carver, Lesser. (1992). "Blackboard Systems for Knowledge-Based Signal Understanding". Oppenheim and Nawab (eds.). "Symbolic and Knowledge-Based Signal Processing". Prentice Hall.

No context found.

N. Carver and V. Lesser. Blackboard systems for knowledge-based signal understanding. In Symbolic and Knowledge-based Signal Processing, pages 205#250, A. V. Oppenheim and S. H. Nawab #eds.#, Prentice-Hall, NJ, 1992.

No context found.

N Carver, V Lesser (1992) "Blackboard systems for knowledge-based signal understanding," in Symbolic and knowledge-based signal processing, ed. AV Oppenheim & SH Nawab, Prentice Hall

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC