Abstract We introduce a framework of multiple viewpoint systems for describing and designing systems that use more than one representation or set of relevance judgments on the same collection. A viewpoint is any representational scheme on some collection of data objects together with a mechanism for accessing this content. A multiple viewpoint system allows a searcher to pose queries to one viewpoint and then change to another viewpoint while retaining a sense of context. Multiple viewpoint systems are well suited to alleviate vocabulary mismatches and to take advantage of the possibility of combining evidence. We discuss some of the issues that arise in designing and using such systems and illustrate the concepts with several examples.

This paper presents a model of the automated planning problem that can extend the range of tractable problems, and decrease the solution time for already tractable ones. It uses the concept of midpoint navigation to reduce planning problems with solutions of length N, into sub-problems with reduced length. Due to the exponential elaboration time of planning problems, this provides significant benefit to the planner. However, a critical aspect of this approach is selecting 'good' midpoints for evaluation. We argue that several established data mining techniques might be used to determine these waypoints, and outline methods for applying these techniques.

We study a class of steady-state genetic algorithms where, at each time step, two parents are selected to produce a child which then replaces one member of the population at the next time step. We consider the finite-population case. A general crossover and mutation operation are defined, as well as a genomic distance between individuals. Certain specific properties are required to hold for such operations and distance functions, and we present examples of crossover operations, mutation operations, and distance functions which meet the requirements. We then define the sum over all pairwise population distances as a measure of the diversity of a population and consider the time evolution of the expected diversity of a population. We show conditions where, under uniform, independent selection of parents and the individual to be replaced, the expected diversity monotonically approaches a fixed point. For this case we calculate an explicit formula for the expected diversity at each time step.

Autonomous robots must be designed to function in real-world domains. These domains are characterized by uncertainty, yet little work has been done to quantify the effects of uncertainty on goal satisfaction. Our research on simulated repair robots indicates that different types of uncertainty have fundamentally different relationships to plan success. This paper investigates the impacts of sensor inaccuracies, exogenous events and failures in both motion and gripper operations. Extensive simulations provide the basis for mathematical models of the relationship between these types of uncertainties and plan success. In addition, the impact of current heuristics such as ‘retry on failure’, and ‘tell me three times’ is investigated. Concluding that even low levels of uncertainty can significantly impact goal satisfaction in a stimulated repair robot domain and that repeated attempts provide significant improvement in plan success rates in the face of uncertainty.

We present the concept of "problem perturbation", give the specifications of an application to the travelling salesperson problem (TSP), then discuss an empirical study of the fitness landscape induced by the variation operators in that application.

We consider the problem of estimating the states of a static set of targets, given a collection of densities, each representing the state of a single target. We assume there is no a priori knowledge of which of the given densities represent common targets, but that a prior density for the target locations is available. For a two-dimensional (2-D) location estimation problem, we construct a prior density model based on known features of the terrain. We then give a simple Gaussian association-estimation algorithm using the prior density and present some simulation results. We briefly discuss extensions to nonstatic models.

Motion smear is an important visual cue for motion perception by the human vision system (HVS). However, in image analysis research, exploiting motion smear has been largely ignored. Rather, motion smear is usually considered as a degradation of images that needs to be removed. In this paper, the authors establish a computational model that estimates image motion from motion smear information-"motion from smear". In many real situations, the shutter of the sensing camera must be kept open long enough to produce images of adequate signal-to-noise ratio (SNR), resulting in significant motion smear in images. The authors present a new motion blur model and an algorithm that enables unique estimation of image motion. A prototype sensor system that exploits the new motion blur model has been built to acquire data for "motion-from-smear". Experimental results on images with both simulated smear and real smear, using the authors' "motion-from-smear" algorithm as well as a conventional motion estimation technique, are provided. The authors also show that temporal aliasing does not affect "motion-from-smear" to the same degree as it does algorithms that use displacement as a cue. "Motion-from-smear" provides an additional tool for motion estimation and effectively complements the existing techniques when apparent motion smear is present.

Image motion estimation usually relies on using interframe displacement as the information cue for motion. Another cue has been proposed, namely motion smear. An algorithm used to compute the image motion from the smear information ("motion-from-smear") was proposed by Chen, Nandhakumar and Martin (see Proc. IEEE Intern. Conf. Image Proc., vol.III, p.761-765, Austin, Texas, November 1994) where the algorithm was verified with simulated motion smeared images. This paper extends the work of Chen et al. by presenting (1) experimental results on real data, (2) a feasible sensor design and (3) a generalization to space-varying motion.

In binocular visual systems, vergence is the process of directing the gaze so that the optical axes intersect at a surface point. Correlation-based methods of disparity analysis provide fast estimates of the vergence error. Unfortunately most correlation techniques do not provide mechanisms to determine which image locations contributed to a given correlation peak. The result is that large correlation peaks may have contributions from image arena not relevant to the vergence task. This paper presents a vergence system that applies a cepstral filter to multiscale images obtained from a dominant-eye binocular sensor. As used by this system, the cepstral filter has two main advantages: it enhances targets through narrow-band signal suppression, and it supports a back-projection operation to determine the image locations associated with particular correlation peaks. The use of multiscale images allows the system to have both high resolution for precision in the final vergence and a large field of view for a wide range of initial camera orientations without undue computational cost.>

Motion smear is an important visual cue for the human vision system (HVS) to perceive motion. However, in image analysis research, using motion smear as a valuable visual cue for motion estimation has remained largely ignored. Rather, motion smear is usually considered as a degradation on images that need to be removed. A computational model has been established to estimate image motion from motion smear information. Both theoretical derivation and experimental results are provided.>

The authors describe experiments using a genetic algorithm for feature selection in the context of neural network classifiers, specifically, counterpropagation networks. They present the novel techniques used in the application of genetic algorithms. First, the genetic algorithm is configured to use an approximate evaluation in order to reduce significantly the computation required. In particular, though the desired classifiers are counterpropagation networks, they use a nearest-neighbor classifier to evaluate features sets and show that the features selected by this method are effective in the context of counterpropagation networks. Second, a method called the training set sampling in which only a portion of the training set is used on any given evaluation, is proposed. Computational savings can be made using this method, i.e., evaluations can be made over an order of magnitude faster. This method selects feature sets that are as good as and occasionally better for counterpropagation than those chosen by an evaluation that uses the entire training set.

Designing a VLSI floorplan calls for arranging a given set of modules in the plane to minimize the weighted sum of area and wire-length measures. A method of solving the floorplan design problem using distributed genetic algorithms is presented. Distributed genetic algorithms, based on the paleontological theory of punctuated equilibria, offer a conceptual modification to the traditional genetic algorithms. Experimental results on several problem instances demonstrate the efficacy of this method and indicate the advantages of this method over other methods, such as simulated annealing. The method has performed better than the simulated annealing approach, both in terms of the average cost of the solutions found and the best-found solution, in almost all the problem instances tried.>

In this paper we are concerned with a time varying set of entities located in a fixed field. These entities are sensed at discrete time instances with a single sensor modality. At a given time instant a collection of bivariate Gaussian sensor reports is produced, estimating the locations of a subset of the entities present in the field. A database of reports is maintained which should ideally contain exactly one report for each entity that has been sensed. Whenever a collection of sensor reports is received the database must be updated to reflect the new information. This updating requires correspondence processing between the database reports and the new sensor reports to determine which pairs of sensor and database reports correspond to the same entity. We present an algorithm for performing this correspondence processing under the assumptions that each new collection of sensor reports contains at most one report for any entity, and that the database can be reasonably assumed to contain at most one report for any entity. This algorithm is based on pairwise distance measures between Gaussian distributions. We consider several such distance measures, and present simulation results indicating that the divergence is a reasonable choice. A formulation of the divergence between two bivariate Gaussians as a scalar product is given. We describe a neural network implementation of our algorithm, along with a proof that the network will converge and, under certain restrictions, exactly compute our correspondence algorithm. INFORMS Journal on Computing, ISSN 1091-9856, was published as ORSA Journal on Computing from 1989 to 1995 under ISSN 0899-1499.

A description is given of Erica, a computer workstation with a unique user interface. The workstation is equipped with imaging hardware and software, which automatically record a digital portrait of the user's eye. From the features of the current portrait, the interface calculates the approximate location of the user's eye-gaze on the computer screen. The computer then executes commands associated with the menu option currently displayed at this screen location. In this way, the user can interact with the computer, run applications software, and manage peripheral devices-all simply by looking at an appropriate sequence of menu options displayed on the screen. The eye-gaze interface technology, its implementation in Erica, and its application as a prosthetic device are described.>

A distributed algorithm for solving the floorplan design problem, called a genetic algorithm with punctuated equilibria (GAPE), is presented. Implementation details and the results of empirical studies are presented. The experiments have demonstrated that GAPE performs consistently better than a recently published simulated annealing approach, with respect to both the average cost of the solutions found and the best solution found.>

Occluding contours from an image sequence with view-point specifications determine a bounding volume approximating the object generating the contours. The initial creation and continual refinement of the approximation requires a volumetric representation that facilitates modification yet is descriptive of surface detail. The ``volume segment'' representation presented in this paper is one such representation.

In this paper we will address several fundamental concepts in the analysis of dynamic scenes and will describe a system that derives three-dimensional object representations from the multiple views provided by dynamic images. The two fundamental concepts are that of the correspondence problem (1) and occlusion analysis (2). The former involves the process which is necessary to span the discontinuity inherent in image sequence representations of dynamic scenes. A detailed description of a dynamic scene analysis system (3) is presented with special emphasis on the two major goals pursued in the work, namely, to lessen the dependence on feature point measurements in a structure from motion system and to develop a descriptive three-dimensional object representation that is suitable for dynamic scene analysis systems. The results are a structure from occluding contours system and a "volume segment" representation scheme.