Thomas M. Strat

Context-based vision: recognizing objects using information from both 2D and 3D imagery

Results from an ongoing project concerned with recognizing objects in complex scene domains, especially in the domain that includes the natural outdoor world, are described. Traditional machine recognition paradigms assume either that all objects of interest are definable by a relatively small number of explicit shape models or that all objects of interest have characteristic, locally measurable features. The failure of both assumptions has a dramatic impact on the form of an acceptable architecture for an object recognition system. In this work, the use of the contextual information is a central issue, and a system is explicitly designed to identify and use context as an integral part of recognition that eliminates the traditional dependence on stored geometric models and universal image partitioning algorithms. This paradigm combines the results of many simple procedures that analyze monochrome, color, stereo, or 3D range images. Interpreting the results along with relevant contextual knowledge makes it possible to achieve a reliable recognition result, even when using imperfect visual procedures. Initial experimentation with the system on ground-level outdoor imagery has demonstrated competence beyond what is attainable with other vision systems. >

Decision analysis using belief functions

Abstract : A primary motivation for reasoning under uncertainty is to derive decisions in the face of inconclusive evidence. Shafers theory of belief functions, which explicitly represents the under constrained nature of many reasoning problems, lacks a formal procedure for making decisions. Clearly, when sufficient information is not available, no theory can prescribe actions without making additional assumptions. Faced with this situation, some assumption must be made if a clearly superior choice is to emerge. In this paper we offer a probabilistic interpretation of a simple assumption;on that disambiguates decision problems represented with belief functions. We prove that it yields expected values identical to those obtained by a probabilistic analysis that makes the same assumption. We maintain a strict separation between evidence that carries information about a situation and assumptions that may be made for disambiguation of choices. In addition, we show how the decision analysis methodology frequently employed in probabilistic reasoning can be extended for use with belief functions. This generalization of decision analysis allows the use of belief functions within the familiar framework of decision trees.

Understanding evidential reasoning

Abstract We address recent criticisms of evidential reasoning, an approach to the analysis of imprecise and uncertain information that is based on the Dempster-Shafer calculus of evidence. We show that evidential reasoning can be interpreted in terms of classical probability theory and that the Dempster-Shafer calculus of evidence may be considered to be a form of generalized probabilistic reasoning based on the representation of probabilistic ignorance by intervals of possible values. In particular, we emphasize that it is not necessary to resort to nonprobabilistic or subjectivist explanations to justify the validity of the approach. We answer conceptual criticisms of evidential reasoning primarily on the basis of the criticisms confusion between the current state of development of the theory — mainly theoretical limitations in the treatment of conditional information — and its potential usefulness in treating a wide variety of uncertainty analysis problems. Similarly, we indicate that the supposed lack of decision-support schemes of generalized probability approaches is not a theoretical handicap but rather an indication of basic informational shortcomings that is a desirable asset of any formal approximate reasoning approach. We also point to potential shortcomings of the underlying representation scheme to treat probabilistic reasoning problems. We also consider methodological criticisms of the approach, focusing primarily on the alleged counterintuitive nature of Dempsters combination formula, showing that such results are the result of its misapplication. We also address issues of complexity and validity of scope of the calculus of evidence.

The Grasper-CL graph management system

Grasper-CL is a system for manipulating and displaying graphs, and for building graph-based user interfaces for application programs. It is implemented in COMMON LISP and CLIM, and has been proven by use in a number of applications. Grasper-CL includes several advances in graph drawing. It contains a graph abstract datatype plus a comprehensive and novel language of operations on that datatype. The appearance of Grasper-CL graphs can be tailored by a wide variety of shape parameters that allow the application to customize the display of nodes and edges for different domains. Default values for shape parameters can be established at several levels. Grasper-CL employs a toolbox approach to graph layout: the system contains a suite of graph layout algorithms that can be applied individually, or in combination to produce hierarchical graph layouts. The system also contains an interactive graph browser.

Explaining evidential analyses

Abstract One of the most highly touted virtues of knowledge-based expert systems is their ability to construct explanations for their lines of reasoning. However, there is a basic difficulty in generating explanations in expert systems that reason under uncertainty using numeric measures. In particular, systems based on evidential reasoning using the theory of belief functions have lacked all but the most rudimentary facilities for explaining their conclusions. Int this paper we review the process whereby other expert system technologies produce explanations, and present a methodology for augmenting an evidential reasoning system with a versatile explanation facility. The method, which is based on sensitivity analysis, has been implemented, and several examples of its use are described.

Using contextual information to set control parameters of a vision process

This paper presents a novel approach to supervised learning of image-understanding tactics based on the use of contextual information. We use a database to store past experiences. From this database and the context elements computed from the task specifications and the input data, we determine whether or not an algorithm is applicable, and which parameters are suitable for it. The database is continuously updated with information of success or failure of the system.