| Integrating Mission, Robot Localization and Communication Requirements Through Collaboration |
06 MAY 2005 |
8 pages |
| Authors:
Charles L. Ortiz; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
 | To develop a de-centralized system for allocating mobile sensor assets controlled by teams of autonomous air vehicle (AVs) and deployed in spatially complex urban environments. The sensor assets are to support AV missions involving the pursuit of one or more mobile intelligent ground adversaries. Since sensor information is typically uncertain and incomplete, AVs must be able to cooperate to coordinate sensor assignments in the most efficient way possible. The environments ... |
|
| Online Query Relaxation via Bayesian Causal Structures Discovery |
2005 |
7 pages |
| Authors:
Ion Muslea; Thomas J. Lee; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| We introduce a novel algorithm, TOQR, for relaxing failed queries over databases; i.e., over-constrained DNF queries that return an empty result. TOQR uses a small dataset to discover the implicit relationships among the domain attributes, and then it exploits this domain knowledge to relax the failed query. TOQR starts with a relaxed query that does not include any constraint, and it tries to add to it as many as possible ... |
|
| Introduction to SPARK. Version 0.3 |
13 JUL 2004 |
24 pages |
| Authors:
David Morley; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| SPARK is a new agent framework, being developed at the Artificial Intelligence Center of SRI International. Its design has been strongly influenced by its predecessor, PRS, and is based on the same Belief Desire Intention (BDI) model of rationality. The motivations for the development of SPARK include: support for development of large-scale agent applications, principled representation of procedures that will enable validation and automated synthesis, flexibility in the delivery platform ... |
|
| A Personalized Calendar Assistant |
2004 |
7 pages |
| Authors:
Pauline M. Berry; Melinda Gervasio; Tomas E. Uribe; Karen Myers; Ken Nitz; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Many calendar tools have become available to organize, display, and track a user's commitments. However, most people still spend a considerable amount of time personally organizing meetings and managing the constant changes and adjustments that must be made to their schedules. Our goal is to provide the technology necessary to manage an individual individual's calendar. The resulting agent will let the user retain control of decisions when necessary and relinquish ... |
|
| A Group-Oriented Framework for Coalitions |
APR 2002 |
11 pages |
| Authors:
Eric Hsu; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Coalitions exemplify the fundamental challenge of inducing coherent group behavior from individualistic agent structures. The Collective-Agents (CA) framework rejects the distinction between individual and group deliberation, on a functional basis. Acknowledging that a group does not think using a brain, and an individual brain is not divisible into multiple minds, the CA framework nevertheless seeks an analogous correspondence between the intentional attitudes of individuals and groups alike. Resulting agents are ... |
|
| Stressed and Unstressed Pronouns: Complementary Preferences |
27 AUG 1996 |
23 pages |
| Authors:
Megumi Kameyama; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| I present a unified account of interpretation preferences of stressed and unstressed pronouns in discourse. The central intuition is the Complementary Preference Hypothesis that predicts the interpretation preference of a stressed pronoun from that of an unstressed pronoun in the same discourse position. The base preference must be computed in a total pragmatics module including commonsense preferences. The focus constraint in Rooth's theory of semantic focus is interpreted to be ... |
|
| Using 3-Dimensional Meshes to Combine Image-Based and Geometry Constraints |
24 AUG 1994 |
31 pages |
| Authors:
Pascal V. Fua; Yvan G. Leclerc; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| A unified framework for 3-D shape reconstruction allows us to combine image-based and geometry-based information sources. The image information is akin to stereo and shape-from-shading, while the geometric information may be provided in the form of 3-D points, 3-D features or 2-D silhouettes. A formal integration framework is critical in recovering complicated surfaces because the information from a single source is often insufficient to provide a unique answer. Our approach ... |
|
| A Storage System for Scalable Knowledge Representation |
23 AUG 1994 |
10 pages |
| Authors:
Peter D. Karp; Suzanne M. Paley; Ira Greenberg; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Twenty years of AI research in knowledge representation has produced frame knowledge representation systems (FRSs) that incorporate a number of important advances. However, FRSs lack two important capabilities that prevent them from scaling up to realistic applications: they cannot provide high-speed access to large knowledge bases (KBs), and they do not support shared, concurrent KB access by multiple users. Our research investigates the hypothesis that one can employ an existing ... |
|
| Registration without Correspondences |
20 AUG 1994 |
19 pages |
| Authors:
Pascal V. Fua; Yvan G. Leclerc; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| In this paper, we present a method for registering images of complex 3-D surfaces that does not require explicit correspondences between features across the images. Our method relies on the use of a full 3-D model of the surface to adjust the position and orientation of the camera by minimizing an objective function based on the projections of the images onto the model. This approach constrains the camera parameters strongly ... |
|
| Advanced Methods of Approximate Reasoning |
31 JUL 94 |
190 pages |
| Authors:
Enrique Ruspini; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| The major goals of the program of research on approximate reasoning of the Artificial Intelligence Center (AIC), SRI International, are: The development of sound formal foundations to explain the different methodologies proposed to solve the problems associated with the processing of imprecise and uncertain information; The identification of criteria to determine the applicability of these methodologies to specific problems; The development of methods for the approximate modeling of real world ... |
|
| Building and Using Scene Representation in Image Understanding |
SEP 1993 |
14 pages |
| Authors:
H. H. Baker; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| The task of having computers able to understand their environments through direct imaging has proved to be formidable. With its beginnings about 30 years ago (1). the field of computer vision has grown as a major part of the pursuit for artificial intelligence. Most elements of this pursuit language understanding, reasoning and planning, speech - are very difficult challenges, but vision, with its high dimensionality of space, time, scale, color, ... |
|
| A Common Knowledge Representation for Plan Generation and Reactive Execution |
09 JUN 1993 |
51 pages |
| Authors:
Sr Wilkins David E.; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| This paper describes the ACT formalism, which is designed to encode the knowledge required to support both the generation of complex plans and reactive execution of those plans in dynamic environments. ACT is a heuristically adequate representation that is useful in practical applications. It serves as an interlingua for Artificial Intelligence technologies in planning and reactive control. The design of the formalism is discussed and its use in practical applications ... |
|
| A Multivalued Logic Approach to Integrating Planning and Control |
JUN 1993 |
97 pages |
| Authors:
Alessandro Saffiotti; Sr Konoliage Kurt G.; Enrique H. Ruspini; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Intelligent agents embedded in a dynamic, uncertain environment should incorporate capabilities for both planned and reactive behavior. Many current solutions to this dual need focus on one aspect, and treat the other one as secondary. We propose an approach for integrating planning and control based on control structures, which link physical movements to abstract action descriptions. Control structures induce behaviors of an agent, expressed as trajectories of control actions in ... |
|
| A Fuzzy Controller for Flakey, An Autonomous Mobile Robot |
MAR 1993 |
36 pages |
| Authors:
Alessandro Saffiotti; Sr. Ruspini Enrique H.; Sr. Konolige Kurt G.; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Controlling the movement of an autonomous mobile robot In real-world unstructured environments requires the ability to pursue strategic goals under conditions of uncertainty, incompleteness, and imprecision. We describe a fuzzy controller for a mobile robot that can take multiple strategic goals into consideration. Through the use of fuzzy logic, goal-oriented behavior (e.g., trying to reach a given location) and reactive behavior (e.g., avoiding previously unknown obstacles on the way) are ... |
|
| A System for Labeling Self-Repairs in Speech |
22 FEB 1993 |
10 pages |
| Authors:
John Bear; John Dowding; Elizabeth Shriberg; Patti Price; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| This document outlines a system for labeling self-repairs in spontaneous speech. The system marks the location and extent of a repair, as well as relevant words in the region of the repair. Together these labels determine the relationship between the "error" and the hypothesized "correction." The system is designed to be able to capture distinctions among different repair patterns while remaining easy to learn, apply, and integrate into existing transcription ... |
|
| The Grasper-CL (Trademark) Graph Management System |
20 JAN 1993 |
36 pages |
| Authors:
Peter D. Karp; John D. Lowrance; Strat Sr. Thomas M.; David E. Wilkins Sr.; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Graphs are virtually ubiquitous in programming applications. Moreover, graph-structured information is especially prevalent in AI applications, and in the COMMON LISP system itself. We can enhance programs that manipulate graph-structured information by providing these programs with graphical user interfaces that draw graphs, and that allow users to interact with drawings of graph nodes and edges. Therefore, it follows that a programming tool that supports the construction of graph-based user interfaces ... |
|
| Distributed Reasoning and Planning |
12 OCT 92 |
|
| Authors:
Kurt G. Konolige Sr; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
|
| Overview of the SRI Cartographic Modeling Environment |
JAN 1992 |
18 pages |
| Authors:
Andrew J. Hanson; Lynn Quam; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| The SRI Cartographic Modeling Environment has been created to support research on interactive, semiautomated, and automated computer-based cartographic activities. The underlying image manipulation capabilities are provided by the SRI ImagCalc(TM) system. The cartographic features and data that can be entered include multiple images, camera models, digital terrain elevation data, point, line, and area cartographic features, and a wide assortment of three-dimensional objects. Interactive capabilities include free-hand feature lighting entry, altering ... |
|
| Interpretation as Abduction |
92 |
68 pages |
| Authors:
Jerry R. Hobbs; Mark Stickel; Douglas Appelt; Paul Martin; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Abduction is inference to the best explanation. In the TACITUS project at SRI we have developed an approach to abductive inference, called weighted abduction , that has resulted in a significant simplification of how the problem of interpreting texts is conceptualized. The interpretation of a text is the minimal explanation of why the text would be true. More precisely, to interpret a text, one must prove the logical form of ... |
|
| A Unified Abductive Treatment of the Intentional and Informational Aspects of Discourse Interpretation: A Preliminary Report |
92 |
16 pages |
| Authors:
Jerry R. Hobbs; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| The problem of interpreting discourse has been subsumed under the general problem faced by intelligent agents of interpreting the situation they are in by explaining the observable facts. The possibility of interpreting an event as the saying by an intelligent agent of a meaningful stretch of discourse is given by an axiom-axiom (18). The ways in which a stretch of discourse can be analyzed into its parts are given by ... |
|
| An Evolutionary Approach to Designing Neural Networks |
OCT 91 |
|
| Authors:
Aviv Bergman; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
 | One of the most interesting properties of neural networks is their ability to learn appropriate behavior by being trained on examples. Established learning algorithms which typically work by minimizing error through backpropagation in weight space, tend to get stuck in local optima--a tendency typical of gradient-descent methods applied to nonconvex objectives functions. Therefore, for problems of nontrivial complexity these systems must be handcrafted to a significant degree, but, the distributed ... |
|
| The TACITUS System: The MUC-3 Experience |
OCT 1991 |
29 pages |
| Authors:
Jerry R. Hobbs; Douglas E. Appelt; John S. Bear; Mabry Tyson; David Magerman; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| SRI International has been engaged in research on text understanding for a number of years. The Naval Ocean Systems Center (NOSC) has sponsored three workshops in recent years for evaluating text understanding systems. SRI participated in the first Message Understanding Conference (MUC-1) in June 1987 as an observer, and subsequently as a participant. Our system was evaluated in the second and third workshops, MUC-2 and MUC-3. For MUC-2, the task ... |
|
| Quantification in Autoepistemic Logic |
06 SEP 1991 |
48 pages |
| Authors:
Kurt Konolige; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Quantification in modal logic is interesting from a technical and philosophical standpoint. Here we look at quantification in autoepistemic logic, which is a modal logic of self-knowledge. We propose several different semantics, all based on the idea that having beliefs about an individual amounts to having a belief using a certain type of name for the individual. |
|
| Resolution for Epistemic Logics |
25 AUG 1991 |
33 pages |
| Authors:
Kurt Konolige; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Quantified modal logics have emerged as useful tools in computer science for reasoning about knowledge and belief of agents and systems. An important class of these logics have a possible-world semantics from Kripke. In this paper we report on a resolution proof method for logics of belief that is suitable for automatic reasoning in commonsense domains. This method is distinguished by its use of an unrestricted first-order modal language, a ... |
|
| Hierarchic Autoepistemic Theories for Nonmonotonic Reasoning: Preliminary Report |
APR 1991 |
28 pages |
| Authors:
Kurt Konolige; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Nonmonotonic logics are meant to be a formalization of nonmonotonic reasoning. However, for the most part they fail to embody two of the most important aspects of such reasoning: the explicit computational nature of nonmonotonic inference, and the assignment of preferences among competing inferences. We propose a method of nonmonotonic reasoning in which the notion of inference from specific bodies of evidence plays a fundamental role. The formalization is based ... |
|
| Abduction vs. Closure in Causal Theories |
APR 1991 |
27 pages |
| Authors:
Kurt Konolige; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| There are two distinct formalizations for reasoning from observations to explanations, as in diagnostic tasks. The consistency based approach treats the task as a deductive one, in which the explanation is deduced from a background theory and a minimal set of abnormalities. In the other treatment, based on abduction, the explanations are considered to be sentences that, when added to the background theory, account for the observations. We show that ... |
|
| Understanding Evidential Reasoning |
DEC 1990 |
25 pages |
| Authors:
Sr. Ruspini Enrique H.; John D. Lowrance; Sr Strat Thomas M.; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| 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. ... |
|
| Proving Properties of Rule-Based Systems |
DEC 1990 |
30 pages |
| Authors:
Richard J. Waldinger; Mark E. Stickel; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Rule-based systems are being applied to tasks of increasing responsibility. Deductive methods are being applied to their validation, to detect flaws in these systems and enable us to use them with more confidence. Each system of rules is encoded as a set of axioms that define the system theory. The operation of the rule language and information about the subject domain are also described in the system theory. Validation tasks, ... |
|
| An Evolutionary Approach to Designing Neural Networks |
SEP 90 |
|
| Authors:
Stephen Barnard; Aviv Bergman; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| One of the most interesting properties of neural networks is their ability to learn appropriate behavior by being trained on examples. Established learning algorithms, which typically work by minimizing error through backpropagation in weight space, tend to get stuck in local optima--a tendency typical of gradient-descent methods applied to nonconvex objective functions. Therefore, for problems of nontrivial complexity these systems must be handcrafted to a significant degree, but the distributed ... |
|
| Approximate Reasoning: Past, Present, Future |
27 JUN 1990 |
26 pages |
| Authors:
Enrique H. Ruspini; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| This note presents a personal view of the state of the art in the representation and manipulation of imprecise and uncertain information by automated processing systems. To contrast their objectives and characteristics with the sound deductive procedures of classical logic, methodologies developed for that purpose are usually described as relying on Approximate Reasoning. Using a unified descriptive framework, we will argue that, far from being mere approximations of logically correct ... |
|
| Translation by Abduction |
01 MAY 1990 |
16 pages |
| Authors:
Jerry R. Hobbs; Megumi Kameyama; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Many existing approaches to machine translation take for granted that the information presented in the output is found somewhere in the input, and, moreover, that such information should be expressed at a single representational level, for example, in terms of the parse trees or of "semantic" assertions. Languages, however, not only express the equivalent information by drastically different linguistic means, but also often disagree in what distinctions should be expressed ... |
|
| Incremental Interpretation |
MAY 1990 |
69 pages |
| Authors:
Fernando C. Pereira; Martha E. Pollack; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| The authors present a system for the incremental interpretation of natural-language utterances in context. The main goal of the work is to account for the influences of context on interpretation, while preserving compositionality to the extent possible. To achieve this goal, they introduce a representational device, conditional interpretations, and a rule system for constructing them. Conditional interpretations represent the potential contributions of phrases to the interpretation of an utterance. The ... |
|
| Introducing the Tileworld: Experimentally Evaluating Agent Architectures |
MAY 1990 |
9 pages |
| Authors:
Martha E. Pollack; Marc Ringuette; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| We describe a system called Tileworld, which consists of a simulated robot agent and a simulated environment which is both dynamic and unpredictable. Both the agent and the environment are highly parameterized, enabling one to control certain characteristics of each. We can thus experimentally investigate the behavior of various meta-level reasoning strategies by tuning the parameters of the agent, and can assess the success of alternative strategies in different environments ... |
|
| Planning and Imperative Program Synthesis: A Deductive Approach |
26 MAR 90 |
|
| Authors:
Richard Waldinger; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| A theorem proving framework is developed particularly well suited to the program synthesis application. In this framework, a deductive tableau of assertions and goals is manipulated, with declarative sentences each associated with a term, called its output entry. While the assertions and goals of the tableau are all that we need for a pure theorem-proving task, the output entries are required for extracting a program from a proof. The deduction ... |
|
| Overloading Intentions for Efficient Practical Reasoning |
1990 |
36 pages |
| Authors:
Martha E. Pollack; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Agents, whether biological or artificial, have bounded reasoning capabilities. As a result, they cannot make reasoned decision instantaneously; reasoning takes time. Agents in dynamic environments face a potential difficulty when they must make decisions what to do. They run the risk the world may change in ways that undermine the very assumptions upon which their reasoning is proceeding. Dynamic environments and computational resource bounds thus pose a challenge that has ... |
|
| The Path-Indexing Method for Indexing Terms |
OCT 1989 |
28 pages |
| Authors:
Mark E. Stickel; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| The path-indexing method for indexing first-order predicate calculus terms is a refinement of the standard coordinate-indexing method. Path indexing offers much faster retrieval at a modest cost in space. Path indexing is compared with discrimination-net and codeword indexing. While discrimination-net indexing may often be the preferred method for maximum speed, path indexing is an effective alternative if discrimination-net indexing requires too much space or in certain cases in which discrimination-net ... |
|
| Decision Analysis Using Belief Functions |
27 SEP 1989 |
34 pages |
| Authors:
Thomas M. Strat; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| A primary motivation for reasoning under uncertainty is to derive decisions in the face of inconclusive evidence. Shafer's 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 ... |
|
| Fast Parallel Surface Interpolation With Applications to Digital Cartography |
16 JUN 1989 |
41 pages |
| Authors:
Richard Szeliski; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| The manipulation of two dimensional elevation maps is an important part of digital cartography. In many situations, these maps are computed by interpolating sparse data such as isolated elevation points obtained from stereo matching. In this paper, we present a surface interpolation algorithm based on variational splines which is well suited to massively parallel computers. Using multiresolution parallel relaxation, we can efficiently compute the interpolated surface and also have local ... |
|
| Objective Functions for Feature Discrimination |
MAY 1989 |
9 pages |
| Authors:
Pascal Fua; Andrew J. Hanson; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| We propose and evaluate a class of objective functions that rank hypotheses for feature labels. Our approach takes into account the representation cost and quality of the shapes themselves, and balances the geometric requirements against the photometric evidence. This balance is essential for any system using Under constrained or generic feature models. We introduce examples of specific models allowing the actual computation of the terms in the objective function, and ... |
|
| Morphology with Two-Level Rules and Negative Rule Features |
20 MAR 1989 |
9 pages |
| Authors:
John Bear; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Two-level phonology, as currently practiced, has two severe limitations. One is that phonological generalizations are generally expressed in terms of transition tables of finite-state automata, and these tables are cumbersome to develop and refine. The other is that lexical idiosyncrasy is encoded by introducing arbitrary diacritics into the spelling of a morpheme. This paper explains how phonological rules may be employed instead of transition tables and describes a more elegant ... |
|
| Recognizing Objects in a Natural Environment: A Contextual Vision System (CVS) |
MAR 1989 |
25 pages |
| Authors:
Martin A. Fischler; Thomas M. Strat; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| Existing machine vision techniques are not competent to reliably recognize objects in unconstrained views of natural scenes. In this paper we identify a number of weaknesses in current recognition systems, including an inability to solve the partitioning problem or to effectively use context and other types of knowledge beyond that of immediate object appearance. We propose specific mechanisms for dealing with some of these problems and describe the design of ... |
|
| The Vision Problem: Exploiting Parallel Computation |
28 FEB 1989 |
65 pages |
| Authors:
Martin A. Fischler; Oscar Firschein; Stephen T. Barnard; Pascal V. Fua; Yvan Leclerc; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| This technical report consists of an introductory paper and three technical papers presented at the session, "AI Application of Supercomputers: The Vision Problem," at the Fourth International Conference on Supercomputing, Santa Clara, California, April 30 to May 5, 1989. |
|
| Reimplementation of the Stanford Stereo System. Integration Experiments with the SRI Baseline Stereo System |
FEB 1989 |
14 pages |
| Authors:
H. H. Baker; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| We describe experiments in stereo matching using a Lisp Machine implementation of the Baker stereo system developed at Stanford University. The processing is one of edge matching in a hierarchy of long to short image contours, finishing with interedge intensity correlation to yield a dense map of scene disparities. An experiment and the results obtained in coupling this with the SRI STEREOSYS mapping system are presented. |
|
| An Integrated Framework for Semantic and Pragmatic Interpretation |
30 JAN 1989 |
14 pages |
| Authors:
Martha E. Pollack; Fernando C. Pereira; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| We report on a mechanism for semantic and pragmatic interpretation that has been designed to take advantage of the generally compositional nature of semantic analysis without unduly constraining the order in which pragmatic decisions are made. To achieve this goal we introduce the idea of a conditional interpretation: one that depends upon a set of assumptions about subsequent pragmatic processing. Conditional interpretations are constructed compositionally according to a set of ... |
|
| P-PATR: A Compiler for Unification-Based Grammars |
15 SEP 1988 |
104 pages |
| Authors:
Susan B. Hirsh; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| P-PATR is a compiler for unification-based grammars that is written in Quintus Prolog running on a Sun 2 workstation. P-PATR is based on the PATR-II formalism [14] developed at SRI International. PATR is a simple, unification-based formalism capable of encoding a wide variety of grammars. As a result of this versatility, several parsing systems and development environments based on this formalism have been implemented [18,5]. P-PATR is one such system, ... |
|
| A Prolog-like Inference System for Computing Minimum-Cost Abductive Explanations in Natural-Language Interpretation |
SEP 1988 |
23 pages |
| Authors:
Mark E. Stickel; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| By determining what added assumptions would suffice to make the logical form of a sentence in natural language provable, abductive inference can be used in the interpretation of sentences to determine what information should be added to the listener's knowledge, i.e., what he should learn from the sentence. This is a comparatively new application of mechanized abduction. A new form of abduction-least specific abduction-is proposed as being more appropriate to ... |
|
| Rex Programmer's Manual |
01 JUL 1988 |
44 pages |
| Authors:
Leslie Pack Kaelbling; Nathan J. Wilson; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| This manual describes Rex, a programming language for specifying machines by dedara tively describing their behavior. The Rex language consists of a set of LISP functions that define primitive Rex machines and provides methods for building complex machines out of simpler components. A Rex machine is a synchronous abstract device that has inputs, local state, and outputs, all of which are storage locations. Storage locations may be thought of as ... |
|
| Classification-Based Tracking of Objects and Materials |
JUL 1988 |
26 pages |
| Authors:
Kenneth I. Laws; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| SRI's KNIFE image analysis system can be used for tracking objects and material classes from one image to another. Variations on this theme are the initial acquisition of target instances from database signatures and the subsequent acquisition of additional instances in an image once a few objects have been labeled. Classification-based tracking is facilitated by improved color and texture-energy transforms. KNIFE's labeling and partitioning methods can be used with complex ... |
|
| Coarse Coding for Material and Object Identification |
JUL 1988 |
28 pages |
| Authors:
Kenneth I. Laws; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| A new coarse-coding technique is presented for labeling image pixels and regions to match exemplars or multivariate material signatures. This multinomial classification method can be used for object cueing and tracking, as well as for material identification and image segmentation. Pixels are classified--and classification reliability can be estimated--with only single-band histograms and one pass through each image band. An example of four-class labeling illustrates the power of this two-level classification ... |
|
| A Survey of Architectures for Distributed Artificial Intelligence |
JUN 1988 |
44 pages |
| Authors:
Todd R. Davies; SRI INTERNATIONAL MENLO PARK CA ARTIFICIAL INTELLIGENCE CENTER
|
| This report surveys literature and research in the field of distributed artificial intelligence (DAI) and provides an overview of computer architectures particularly suited to such research. It concentrates on work to date that has involved the construction of testbeds and development tools for DAI. It tries to draw some lessons from these efforts and suggests ways in which testbeds, which heretofor have been used primarily for experimentation, might be used ... |
|
Reports by Corporate Author
