leff@smu.UUCP (Laurence Leff) (12/29/89)
NEW MIT AI LAB PUBLICATIONS Announcement #14 December 28, 1989 ************************************************************************ Except for the Bibliographies, publications are available in hardcopy only. TO ORDER, specify publications number and author and enclose a check payable to the M.I.T. Artificial Intelligence Laboratory for the correct amount of U.S. funds. Prices of publications include surface postage to domestic and overseas addresses. PREPAYMENT IS REQUIRED. Please note on order if check is sent separately. Send orders with payment to: Publications, Room NE43-818 M.I.T. Artificial Intelligence Laboratory 545 Technology Square Cambridge, MA 02139 USA For additional information: Phone number: (617) 253-6773 Net address: Publications@ai.mit.edu *********************************************************************** :aim 1146 :author Shimon Edelman and Daphna Weinshall :asort Edelman, S.; Weinshall, D. :title A self-organizing multiple-view representation of three-dimensional objects :date August 1989 :pages 27 :cost $3.25 :keywords recognition, network models, learning, representation :abstract We explore representation of 3D objects in which several distinct 2D views are stored for each object. We demonstrate the ability of a two-layer network of threshold summation units to support such representations. Using unsupervised Hebbian relaxation, we trained the network to recognize ten objects from different viewpoints. The training process led to the emergence of compact representations of the specific input view. When tested on novel views of the same objects, the network exhibited a substantial generalization capability. In simulated psychophysical experiments, the network's behaviour was qualitatively similar to that of human subjects. :aim 1148 :author Anita M. Flynn and Rodney A. Brooks :asort Flynn, A.; Brooks, R. :title Battling Reality :date October 1989 :pages 21 :cost $3.25 :keywords subsumption architecture, mobile robots, gnat robots, robotics, sensors, navigation :abstract In the four years that the MIT Mobile Robot Project has been in existence, we have built ten robots that focus research in various areas concerned with building intelligent systems. Towards this end, we have embarked on trying to build useful autonomous creatures that live and work in the real world. Many of the preconceived notions entertained before we started building our robots turned out to be misguided. This paper describes the changing paths our research has taken due to the lessons learned from the practical realities of building robots. :aim 1157 :author Thomas Marill :asort Marill, T. :title Recognizing Three-Dimensional Objects Without The Use of Models :date September 1989 :pages 26 :cost $3.25 :keywords recognition, vision, three-dimensional, line drawings :abstract We present an approach to the problem of recognizing three-dimensional objects from line-drawings. In this approach there are no models. The system needs only to be given a single picture of an object; it can then recognize the object in arbitrary orientations. :aim 1164 :author Federico Girosi and Tomaso Poggio :asort Girosi, F.; Poggio, T. :title Networks and the Best Approximation Property :date October 1989 :pages 22 :cost $3.25 :keywords learning, networks, regularization, best approximation, approximation theory :abstract Networks can be considered as approximation schemes. Multilayer networks of the backpropagation type can approximate arbitrarily well continuous (Cybenko, 1989; Funahashi, 1989; Stinchcombe and White, 1989). We prove that networks derived from regularization theory and including Radial Basis Function (Poggio and Girosi, 1989), have a similar property. From the point of view of approximation theory, however, the property of approximating continous functions arbitrarily well is not sufficient for characterizing good approximation schemes. More critical is the property of best approximation. The main result of this paper is that multilayer networks, of the type used in backpropagation, are not best approximation. For regularization networks (in particular Radial Basis Function networks) we prove existence and uniqueness of best approximation. :aim 1165 :author Brian LaMacchia and Jason Nieh :asort LaMacchia, B.; Nieh, J. :title The Standard Map Machine :date September 1989 :pages 47 :cost $3.75 :keywords chaos, nonlinear mappings, numerical computations, computer architecture, special-purpose computing :abstract We have designed the Standard Map Machine(SMM) as an answer to theintensive computational requirements involved in the study of chaotic behavior in nonlinear systems. The high-speed and high-precision performance of this computer is due to its simple architecture specialized to the numerical computations required of nonlinear systems. In this report, we discuss the design and implementation of this special-purpose machine. :aim 1174 :author Harold Abelson :asort Abelson, H. :title The Bifurcation Interpreter: A step towards the automatic analysis of dynamical systems :date September 1989 :cost $3.75 :adnum AD-A214639 :pages 43 :keywords dynamical systems, algebraic/symbolic computation, numeric/symbolic processing, scientific computing :abstract :aim 1176 :author David McAllester and Robert Givan :asort McAllester, D.; Givan, R. :title Natural Language Syntax and First Order Inference :date October 1989 :cost $3.25 :pages 28 :keywords knowledge representation, automated reasoning, Montague semantics, theorem proving, natural language, taxonomy :abstract We have argued elsewhere that first order inference can be made more efficient by using non-standard syntax for first order logic. In this paper we show how a fragment of English syntax under Montague semantics provides the foundation of a new inference procedure. This procedure seems more effective than corresponding procedures based on either classical syntax or our previously proposed taxonomic syntax. This observation may provide a functional explanation for some of the syntactic structure of English. :aim 1177 :author David W. Jacobs :asort Jacobs, D. :title Grouping for Recognition :date November 1989 :cost $3.75 :pages 44 :keyword grouping, indexing, recognition, model-based vision :abstract This paper presents a new method of grouping edges in order to recognize objects. This grouping method succeeds on images of both two- and three-dimensional objects. So that the recognition system can consider first the collections of edges most likely to lead to the correct recognition of objects, we order groups of edges based on the likelihood that a single object produced them. The grouping module estimates this likelihood using the distance that separates edges and their relative orientation. This ordering greatly reduces the amount of computation required to locate objects, and improves the systems robustness to error. :aim 1183 :author Andrew Trice and Randall Davis :asort Trice, A.; Davis, R. :title Consensus Knowledge Acquisition :date December 1989 :cost $3.25 :pages 25 :contract N00014-85-K-0124 :keyword knowledge acquisition, multiple experts, knowledge based systems, debugging :abstract We have developed a method and prototype program for assisting two experts in their attempts to construct a single, consensus knowledge base. We show that consensus building can be effectively facilitated by a debugging approach that identifies, explains, and resolves discrepancies in their knowledge. To implement this approach we identify and use recognition and repair procedures for a variety of discrepancies. Examples of this knowledge are illustrated with sample transcripts from {\sc carter}, a system for reconciling two rule-based systems. Implications for resolving other kinds of knowledge representations are also examined. :tr 1113 :author Karen Beth Sarachik :asort Sarachik, K. :title Visual Navigation: Constructing and Utilizing Simple Maps of an Indoor Environment :date March 1989 :cost $8.00 :pages 91 :keywords map making, mobile robot, motion vision, robot navigation, stereo vision :abstract The goal of this work is to navigate through an office environment using only visual information gathered from four cameras placed onboard a mobile robot. The method is insensitive to physical changes within the room it is inspecting, such as moving objects. Forward and rotational motion vision are used to find doors and rooms, and these can be used to build topological maps. The map is built without the use of odometry or trajectory integration. The long term goal of the project described here is for the robot to build simple maps of its environment and to localize itself within this framework. :tr 1161 :author Lyle Borg-Graham :asort Borg-Graham, L. :title Modelling the Somatic Electrical Response of Hippocampal Pyramidal Neurons :date September 1989 :pages 295 :cost $10.00 :keywords computational neuroscience, hippocampus, modelling :abstract A modelling study of hippocampal pyramidal neurons is described. This study is based on simulations using HIPPO, a program which simulates the somatic electrical activity of these cells. HIPPO is based on descriptions of eleven non-linear conductances that have been either reported in the literature or postulated in the present study, and an approximation of the electrotonic structure of the cell that is derived in this thesis. HIPPO is used a) to integrate electrophysiological data from a variety of sources, b) to investigate the functional significance of the elements underlying electrical behavior, and c) to provide a tool for the electrophysiologist to supplement direct observation of these cells and allow testing of speculations regarding unobservable parameters. :tr 1179 :author Marc H. Raibert, H. Benjamin Brown, Jr., Michael Chepponis, Jeff Koechling, Jessica K. Hodgins, Diane Dustman, W. Kevin Brennan, David S. Barrett, Clay M. Thompson, John Daniell Hebert, Woojin Lee, Lance Borvansky :asort Raibert, M.; Brown, H.B.; Chepponis, M.; Koechling, J.; Hodgins, J.; Dustman, D.; Brennan, W.K.; Barrett, D.; Thompson, C.; Hebert, J.; Lee, W.; Borvansky, L. :title Dynamically Stable Legged Locomotion (September 1985-September 1989) :date September 1989 :pages 203 :cost $9.00 :keyword robotics, legged locomotion, dynamic stability, gymnastics, legged robots, robot control :abstract This report documents our work in exploring active balance for dynamic legged systems for the period from September 1985 through September 1988. The purpose of this research is to build a foundation of knowledge that can lead both to the construction of useful legged vehicles and to a better understanding of animal locomotion. In this report we focus on the control of biped locomotion, the use of terrain footholds, running at high speed, biped gymnastics, symmetry in running, and the mechanical design of articulated legs.