## SEARCH

#### Country

##### ( see all 69)

- United States 2299 (%)
- Germany 2232 (%)
- United Kingdom 1327 (%)
- France 722 (%)

#### Institution

##### ( see all 6405)

- Cornell University 127 (%)
- OneVision GmbH 106 (%)
- University of California 94 (%)
- Technische Universität Wien 89 (%)
- Carnegie Mellon University 84 (%)

#### Author

##### ( see all 18513)

- Hamann, Sabine 106 (%)
- Hell, Hauke 106 (%)
- Pankow, Detlef 106 (%)
- Wunderer, Robert 106 (%)
- Baumgardt, Michael 60 (%)

#### Publication

##### ( see all 499)

- Artificial Neural Networks — ICANN'97 204 (%)
- Computer Vision — ACCV'98 196 (%)
- Image Analysis and Processing 192 (%)
- Euro-Par'97 Parallel Processing 188 (%)
- Human-Computer Interaction INTERACT ’97 150 (%)

#### Subject

##### ( see all 175)

- Computer Science [x] 14094 (%)
- Artificial Intelligence (incl. Robotics) 4444 (%)
- Theory of Computation 2620 (%)
- Computer Communication Networks 2561 (%)
- Software Engineering/Programming and Operating Systems 2106 (%)

## CURRENTLY DISPLAYING:

Most articles

Fewest articles

Showing 1 to 10 of 14094 matching Articles
Results per page:

## Back Matter - Algebraic and Logic Programming

### Algebraic and Logic Programming (1997-01-01): 1298 , January 01, 1997

## The design and implementation of a multimedia system for the interactive and time-synchronised analysis of vehicle crash tests

### Multimedia Applications, Services and Techniques — ECMAST '97 (1997-01-01) 1242: 79-94 , January 01, 1997

This paper describes the requirements for, and the development and implementation of, an interactive multimedia system for the visualisation and analysis of the results from automobile impact tests and impact simulations. However, the technology described can be applied to a much wider range of scientific and engineering test scenarios.

Traditionally, crash test results are presented to the customer on a number of individual media such as high speed film reels, bulky text reports and photographs. The system described replaces these disparate media with an electronic report on a single CD-ROM for presentation on a multimedia PC to enable integrative and interactive analysis.

In developing this solution, two complementary systems are required. The first collates all the data and produces a database on CD for each test. This system needs to be used in a commercial environment to rapidly produce the database in a routine manner. This is achieved by the use of intuitive interfaces and extensive automation capability in the generation and population of the multimedia relational database. The second system interrogates the CD and presents the information contained therein in an intuitive and useful manner. A major feature of the presentation system is the playback of multiple video streams simultaneously with Finite Element models and multiple time dependent data graphs, all of which are time synchronised.

The systems have been developed using Rapid Application Development (RAD) tools within an object-oriented framework, with Human Computer Interaction (HCI) issues being to the fore at all stages. It is now in routine use at a major crash laboratory, with at least half of their test results being presented in this manner.

## Tight lower bounds for computing shortest paths on proper interval and bipartite permutation graphs

### Parallel Computing Technologies (1997-01-01) 1277: 7-12 , January 01, 1997

Logarithmic time lower bounds for computing the distance between two arbitrary vertices, in a proper interval graph represented by a family of intervals on a real line, and in a bipartite permutation graph represented by a permutation function, on exclusive write PRAM are proved here. The lower bounds are also valid for these classes of graphs represented by adjacency matrices and for their superclasses. Shortest paths on interval and permutation graphs, which, respectively, strictly contain proper interval and bipartite permutation graphs, are known to be computable in logarithmic time on exclusive write PRAM. It follows that the lower bounds derived here are tight.

## Cross-correlations in sparsely connected recurrent networks of spiking neurons

### Artificial Neural Networks — ICANN'97 (1997-01-01) 1327: 31-36 , January 01, 1997

We study the dynamics of sparsely connected recurrent networks composed of excitatory and inhibitory integrate-and-fire (IF) neurons firing at low rates, and in particular cross-Correlations (CC) between spike times of pairs of neurons using both numerical simulations and a recent theory. CCs exhibit damped oscillations with a frequency which depends on synaptic time constants. Individual CCs are shown to depend weakly on synaptic connectivity. They depend more strongly on the firing rates of individual neurons.

## Neural network based processing for smart sensors arrays

### Artificial Neural Networks — ICANN'97 (1997-01-01) 1327: 565-570 , January 01, 1997

Source separation (SS) algorithm is an attractive approach for designing smart sensor array, able to increase spatial selectivity and to cancel spurious sources. The source number being unknown and able to vary, a pre-processing algorithm is developped in this paper for providing estimation of the source number before source separation. On-line source separation is then achieved in the above time variant context.

## Integration von Bildsynthese und Bildanalyse

### Von der Photographie zum 3D-Modell (1997-01-01): 7-30 , January 01, 1997

### Zusammenfassung

In diesem Kapitel werden die Gemeinsamkeiten, Unterschiede und möglichen Verbindungen von Computer Graphik und Computer Vision untersucht. Zur Motivation der nachfolgenden Ausführungen, werden zu Beginn allgemeine Gesichtspunkte bezüglich einer Integration von Computer Graphik und Computer Vision dargestellt. Anschließend wird die algorithmische Dualität zwischen Bildanalyse und Bildsynthese betrachtet und gleichzeitig werden Möglichkeiten für eine Harmonisierung der Verfahren aufgezeigt. Ein besonderer Augenmerk gilt hierbei den Szenenbeschreibungsattributen Geometrie, Beleuchtung, Textur und Bewegung, da diese sowohl im Bereich der Bildanalyse als auch für die Bildsynthese einen maßgeblichen Stellenwert besitzen. Am Ende des Kapitels erfolgt ein konzeptioneller Entwurf eines Referenzmodells zur Zusammenführung der Bereiche Computer Graphik und Computer Vision. Eine Verfeinerung des Modells wird in Kapitel 5 durchgeführt.

## CICS, IMS And Others Go Web Via Business Objects

### OOIS’96 (1997-01-01): 486-493 , January 01, 1997

CICS and IMS applications can be accessed by WWW using business objects that are generated from CICS or IMS applications. An application programmer builds business objects and fills them with data via a tool that encapsulates transactions. The business objects’ attributes can be connected to a user interface running on the same computer or to a HTML page that allows to access the business objects and thus the data from the underlying transaction system.

## From the Art of KDD to the Science of KDD

### Learning, Networks and Statistics (1997-01-01) 382: 135-160 , January 01, 1997

It has been already largely proven that *Knowledge Discovery in Databases* (KDD) is an interesting new research field, able to provide financial returns to the companies that are willing to invest into it. This fact demonstrates the excellent social value of KDD. A Science, however, is not uniquely defined by this feature. It needs also to show an internal logic, due to a specific approach to the real-life problems it deals with. This last point of view has been less emphasized in the existing KDD literature. This paper attempts, without any pretense to be exhaustive, to start filling up this gap. We shall explain why KDD is not just “a bunch of techniques” but a real Science, certainly one still under organization, but which shows the strong inner motivation that other Sciences do. In conclusion we shall give a compact definition of KDD, and show what is the concept it provides measurement of, as a function of which other concepts.

## Online Learning versus Offline Learning

### Machine Learning (1997-10-01) 29: 45-63 , October 01, 1997

We present an off-line variant of the mistake-bound model of learning. This is an intermediate model between the on-line learning model (Littlestone, 1988, Littlestone, 1989) and the self-directed learning model (Goldman, Rivest & Schapire, 1993, Goldman & Sloan, 1994). Just like in the other two models, a learner in the off-line model has to learn an unknown concept from a sequence of elements of the instance space on which it makes “guess and test” trials. In all models, the aim of the learner is to make as few mistakes as possible. The difference between the models is that, while in the on-line model only the set of possible elements is known, in the off-line model the sequence of elements (i.e., the identity of the elements as well as the order in which they are to be presented) is known to the learner in advance. On the other hand, the learner is weaker than the self-directed learner, which is allowed to choose adaptively the sequence of elements presented to him.

We study some of the fundamental properties of the off-line model. In particular, we compare the number of mistakes made by the off-line learner on certain concept classes to those made by the on-line and self-directed learners. We give bounds on the possible gaps between the various models and show examples that prove that our bounds are tight.

Another contribution of this paper is the extension of the combinatorial tool of labeled trees to a unified approach that captures the various mistake bound measures of all the models discussed. We believe that this tool will prove to be useful for further study of models of incremental learning.