The effective use between these applications of the limited bandwidth at the User-Network Interface (UNI) has raised increasing interest from both end-system and telecommunication networks researchers. This paper addr...
详细信息
The effective use between these applications of the limited bandwidth at the User-Network Interface (UNI) has raised increasing interest from both end-system and telecommunication networks researchers. This paper addresses the structure of higher layer traffic controls at the end-systems. A session scheduling priority scheme is proposed which reflects the user specified service priority and the importance of the minimum quality. A stream level traffic control device ('stream filter') is proposed at the end-system, which may restrain a traffic stream in response to cell level connection status. The interactive mechanism between the session scheduler and the stream filters has been investigated.< >
The problems faced by people whose physical impairments include the inability to communicate through speech are demonstrated by the fact that, even with current technology, speech rates achievable using speech output ...
详细信息
The problems faced by people whose physical impairments include the inability to communicate through speech are demonstrated by the fact that, even with current technology, speech rates achievable using speech output devices typically vary between 2-10 words per minute. We have been investigating ways in which the communication system itself could take on many of the cognitive and manipulative tasks required to locate and produce suitable prestored texts for the user. The storage and retrieval methods we have investigated include a text database with items indexed semantically and pragmatically, logging conversational paths for reuse, hypertext, and fuzzy information retrieval. The ultimate goal of this work is a communication system which will allow a severely impaired nonvocal person to make a significantly better conversational impact than is possible with present systems.< >
We present a cognitive approach to generalisation, Bidirectional Convergence. This is the implementation of a cognitive process we call concept crystallisation, whereby a concept is formed gradually from initially man...
详细信息
We present a cognitive approach to generalisation, Bidirectional Convergence. This is the implementation of a cognitive process we call concept crystallisation, whereby a concept is formed gradually from initially many possibilities which converge to a single possibility under the weight of a series of learning instances shown over a period of time. Bidirectional Convergence (BDC) is a form of concept crystallisation that represents the alternative possible concepts through. Boundary versions of the concept during learning. BDC is an abstraction of Mitchell's symbolic concept learning technique (1982). We describe how BDC is evolved from Mitchell's technique into a form suitable for incorporation into neural networks, BDC is shown to provide a best-fit to given problems.< >
Multimedia application integrates a variety of media, namely, audio, video, images, graphics, text and data. The ATM (Asynchronous Transfer Mode) technique has been recommended by the CCITT as the transport vehicle fo...
详细信息
Multimedia application integrates a variety of media, namely, audio, video, images, graphics, text and data. The ATM (Asynchronous Transfer Mode) technique has been recommended by the CCITT as the transport vehicle for BISDN (Broadband Integrated Service Digital networks) which is designed to provide multimedia traffic services. In this work we study some of the ATM-compatible multiple access MAN (Metropolitan Area Network) protocols, including DQDB (Distributed Queue Dual Bus, also known as IEEE 802.6), CRMA (Cyclic Reservation Multiple Access), DQMA (Distributed Queue Multiple Access), and FDQ (Fair Distributed Queue), and their performance for different kinds of multimedia application, including interactive education, on-line medical information system, and video conferences. We found that all four protocols achieve maximum throughput even at extremely high data rate, which make them suitable for carrying multimedia traffic in high speed networks.
Given an arbitrary telecommunications network N our goal is to find the minimum cost for equipment which will enable N to survive an arbitrary link fault. We consider uni-directional and bi-directional ring technologi...
详细信息
Given an arbitrary telecommunications network N our goal is to find the minimum cost for equipment which will enable N to survive an arbitrary link fault. We consider uni-directional and bi-directional ring technologies. Basically, our goal is to find minimum cost ring covers for any network N, where a ring cover is a set C of rings such that every link in N is covered by (i.e. part of) at least one ring in C. If a network N is augmented with enough equipment to support a given ring cover C, it can respond to a link failure immediately (and automatically) by routing the disrupted traffic through surviving links in the ring that covers the failed link. We describe an efficient algorithm to find a minimum cost ring cover for uni-directional transmission rings under simplifying assumptions. This algorithm offers a useful heuristic for comput.ng low cost ring covers for existing networks and actual cost functions. We also provide efficient heuristics to find nearly minimum cost ring covers for bi-directional transmission rings. We show that certain versions of the bi-directional problem are NP-complete, hence (presumably) no efficient algorithm exists that always finds a minimum cost ring cover. However, our heuristics perform well in practice.
Listeners’ abilities to detect changes in tempo were investigated with two‐ and four‐tone isochronous sequences with interonset intervals (IOI) of 100, 400, 700, and 1000 ms. Separate thresholds were measured for i...
Listeners’ abilities to detect changes in tempo were investigated with two‐ and four‐tone isochronous sequences with interonset intervals (IOI) of 100, 400, 700, and 1000 ms. Separate thresholds were measured for increases and decreases in tempo using an adaptive‐tracking procedure. On each trial a standard pattern was followed by two comparison patterns, one of which was faster or slower than the standard. Listeners judged which comparison pattern was different from the standard. Consistent with previous studies of tempo discrimination, thresholds were found to be lower with four‐tone sequences than with two‐tone sequences, especially at the faster tempos. However, at the fastest tempos, listeners showed greater sensitivity to increases than to decreases in tempo, while the reverse was true at the slower tempos. The crossover point occurred at an IOI between 400 and 700 ms. The findings are consistent with the predictions of an entrainment model [J. D. McAuley, J. Acoust. Soc. Am. 95, 2966 (A) (1994)] in which tempo sensitivity is reflected by the degree to which a system of adaptive osci.lators is entrained by the rhythm of a stimulus pattern. [Work supported by NIMH and NIDCD.]
This correspondence presents a metric for describing line segments. This metric measures how well two line segments can be replaced by a single longer one. This depends for example on collinearity and nearness of the ...
详细信息
This correspondence presents a metric for describing line segments. This metric measures how well two line segments can be replaced by a single longer one. This depends for example on collinearity and nearness of the line segments. The metric is constructed using a new technique using so-called neighborhood functions. The behavior of the metric depends on the neighborhood function chosen. In this correspondence, an appropriate choice for the case of line segments is presented. The quality of the metric is verified by using it in a simple clustering algorithm that groups line segments found by an edge detection algorithm in an image. The fact that the clustering algorithm can detect long linear structures in an image shows that the metric is a good measure for the groupability of line segments.
A recent multipoint Pade approximation technique for linear system reduction is extended to include expansion points at infinity. The method is then seen to be completely general, covering the cases of real, complex, ...
详细信息
A recent multipoint Pade approximation technique for linear system reduction is extended to include expansion points at infinity. The method is then seen to be completely general, covering the cases of real, complex, multiple and infinite points. An example is given to illustrate its application.
Previous results on nonlearnability of visual concepts relied on the assumption that such concepts are represented as sets of pixels [1]. This correspondence uses an approach developed by Haussler [2] to show that und...
详细信息
Previous results on nonlearnability of visual concepts relied on the assumption that such concepts are represented as sets of pixels [1]. This correspondence uses an approach developed by Haussler [2] to show that under an alternative, feature-based representation, recognition is PAC learnable from a feasible number of examples in a distribution-free manner.
Spacetimes admitting a group of (local) projective collineations are considered. In an n-dimensional proper Einstein space it is shown that any vector field xi(i) generating a proper projective collineation (that is o...
Spacetimes admitting a group of (local) projective collineations are considered. In an n-dimensional proper Einstein space it is shown that any vector field xi(i) generating a proper projective collineation (that is one which is not an affine collineation) is the gradient of a scalar field phi (up to the addition of a Killing vector field). Then a four-dimensional Einstein spacetime admitting a proper projective collineation is shown to have constant curvature. For an n-dimensional space of non-zero constant curvature, the scalar field phi satisfies a system of third-order linear differential equations. The complete solution of this system is found in closed form and depends on (n + 1)(n + 2)/2 arbitrary constants. All gradient vector fields xi(i) generating projective collineations are found explicitly and together with the n(n + 1) /2 killing vector fields generate a Lie algebra of dimension n(n + 2).
暂无评论