Robotic Control via Stimulus Response Learning. Andreas Birk. Universitat des Saarlandes, c/o Lehrstuhl Prof. W.J. Paul

Size: px
Start display at page:

Download "Robotic Control via Stimulus Response Learning. Andreas Birk. Universitat des Saarlandes, c/o Lehrstuhl Prof. W.J. Paul"

Transcription

1 Robotic Control via Stimulus Response Learning Andreas Birk Universitat des Saarlandes, c/o Lehrstuhl Prof. W.J. Paul Postfach , Saarbrucken, Germany Abstract Stimulus response learning is a new approach to enable robots to learn an adaptive and robust control-mechanism. Learning in this paradigm is an evolutionary process, leading to a behavioral model of any given environment. The model is build up in a special data-structure as collection of regularities between sensor-data and activation of eectors. This process is guided by two simple and universal i.e., independent of the robot and environment, heuristics for rating the usefulness of basic elements of the model. We present a system learning from scratch to control a robot-gripper in a blocks-world via a camera. The system learns movements in the plane and to manipulate building-blocks. The most challenging experiments were done in noisy real world set-ups, using unprocessed high-resolution images. Keywords: Evolutionary Algorithm, Robotics, Adaptive Behavior, Machine Learning 1 Introduction The ability to learn is one of the most important features of intelligence. It enables cognitive systems to adapt to un-foreseen circumstances and to improve with practise. Though these tasks appear to be very easy to humans, automatic adaption and improvement of robots is still a very open research eld. Therefore, we developed stimulus response learning as a new machine learning approach suited for robotic control. We believe, that cognition can only emerge in systems interacting with a suciently complex environment. Or abusing an ancient Roman slogan \mens sana in corpere sano". Though our interests focus on understanding cognition and building a humanoid articial intelligence in long term, stimulus response learning is an interesting approach in regard to industrial applications. Because a learning control-mechanism allows noisy and changing environments, the execution of more various tasks and low-priced imprecise hardware.

2 2 Learning of Control We wish robots to be able to learn to do any given task in any given environment. So, we must supply our robot with a mechanism for investigating his environment, a suitable storage for his discoveries and a mechanism for applying his knowledge learned. 2.1 How to store Discoveries Our robot makes a discovery when he nds useful regularities between sensor-data and activation of eectors. The meaning of \useful" depends on the task(s) the robot is expected to do and will be discussed later. Discoveries are the basic drive for the evolution of the robot's model of the world, which tells him what to do in a certain situation to achieve a certain result. The world-model is built up in a special data-structure, a dynamic directed graph. The nodes are so-called stimulus response rules (SRR), simple behavioral rules made of tests on sensor-data and actions i.e., activation of eectors. In its most general form a SRR is a TOTE (Fig.1), an abbreviation for test 1, operate, test 2, exit. If test 1 the so-called condition holds, the TOTE can be executed. This means, the operate or action is repeatedly activated until test 2 the so-called feedback holds. After this, the exit or result should be fullled. This test represents the state of the environment after execution of the TOTE and is used for planning. false?? true true Condition Action Feedback Result? T est if the SRR is executable O peration of effectors T est if the action has to be repeated E xit to a goal state Figure 1: The TOTE as the most general form of a SRR In the graph representing the world-model, a directed edge (s 1 ; s 2 ) stands for a possible consecutive execution of s 1 and s 2. This means, the system witnessed frequently that the condition of s 2 holds after execution of s 1. An important point is, that this kind of implication is purely based on statistics. 2.2 The Evolution of the World-Model The robot evolves the directed graph representing his world-model in a step-wise manner. In a time-step t the robot chooses randomly between creation of a new SRR or edge and training of the current world-model G t = (V t ; E t ) with V t is the set of SRRs at time-step t and E t V t V t is the set of edges. In doing so, a SRR-pointer, the so-called standpoint is used to mark the SRR that was executed last. It represents the system's present position in the world.

3 Learning as operations on the graph takes mainly place in direct graph-theoretic neighborhood of the standpoint i.e., on M sp = fs 2 V t j(standpoint; s) 2 E t g, the set of SRRs with an edge leading to from the standpoint. This set is so to say the population of our evolutionary algorithm. Therefore, it is not necessary to process the whole world-model learned so far, but a very small subset to breed new knowledge. A new SRR s is created by using parts of SRRs from Msp. In doing so, the systems focuses on SRRs from Msp with good records according to the soon dened quality-measures. Creation of a new edge is done by searching a SRR s with holding condition in the current world-model G t. If the search is successful, an edge from the standpoint to s is included in G t+1 and s is executed. A training-step is done as follows. A SRR s with holding condition is randomly chosen among M sp and executed. Training is necessary to get the statistical records for the quality-measures. After every system-step, no matter if training or creation, all bad with respect to the quality measures SRRs and edges are deleted. 2.3 The Quality-Measures The quality-measures are designed to rate the usefulness of the robot's discoveries. They are a kind of universal tness-function in the evolution of the world-model. They use two simple concepts reliability and applicability and are based on the fact that a SRR or an edge represents an assumption: The execution of a SRR is successful if afterwards 1.) the condition is not holding i.e., the action has changed the state of the environment, and 2.) the feedback became true before the action was repeated more than constant rep max times, and 3.) the result is holding. An edge e = (s 1 ; s 2 ) is successful if the execution of its target-srr s 2 is successful. The reliability of a SRR s, respectively edge e is dened as number of successful executions of s (or e) per number of executions of s (or e) The applicability of a SRR s, respectively edge e is dened as number of executions of s (or e) per lifetime of s (or e) where lifetime is the number of time-steps of the system since creation of s (or e). The quality of a SRR s, respectively edge e is reliability times applicability: number of successful executions of s (or e) per lifetime of s (or e ) 2.4 Using the World-Model Using a graph as world-model planning becomes very easy. As mentioned before, a SRR-pointer standpoint models the system's current position in the world. It is always set on the last executed SRR. Given a desirable state g of the environment the system can search its world-model for a SRR s where the result of s holds on g. To achieve the goal g the system can execute the SRRs along the shortest path from the standpoint to s.

4 3 Learning Eye-Hand-Coordination 3.1 The Task According to the Swiss psychologist Piaget [3] child development proceeds in stages. The rst one is the sensorimotor stage, characterized by eye-hand-coordination. This can be compared to the task of learning to control a robot-arm via a camera (Fig.2): Camerapicture Robotgripper (red) Robotarm (black) Buildingblock (yellow) Background (black) Figure 2: The set-up A camera looks perpendicularly on a black background with some colored buildingblocks on it. A red robot-gripper, the so-called hand, can be moved by one of the actions north, south, west, or east in one of the four directions of the plane. Furthermore, the hand can grip the building block if its right under the hand and carry it around. It can release the block with the action ungrip. 3.2 The World-Model The world-model for moving the hand is a kind of grid. It consists of SRRs of the following form: \If I see the hand at position P, I can take (hand moving) action A and the hand will be visible at position P 0." Four edges are leading from every SRR except those on the edges of the grid to its neighbors in the four directions of the plane. If the system is told to get the hand to a position P in the current video image, the system searches a SRR s with a result representing P and executes the actions of the SRRs of the shortest path from the standpoint to s. The complete world-model for hand-movements and manipulating building-blocks consists of two such grids. The rst one is the above described \eye-hand-grid". The second one describes how building-blocks can be carried around. The two grid are interconnected by SRRs for proper grasping and un-grasping building-blocks i.e., grip is only used if and only if the block is under the hand, ungrip is only used if and only if the block is held. The hardness of the task to learn this world-model is twofold. First, the system has to learn appropriate representations of its hand and blocks dependent on the kind of tests used. We will present shortly results from experiments with two dierent kind

5 of tests. Second, the system has to nd out what is important to look at. For example the system has to discover that hand-movements are independent of block-positions, but grasping is not. 3.3 Teststrings and Gary L. Drescher In one class of experiments the camera picture is sectioned as a grid. In each eld of the grid the most frequent color in it is determined and written to a vector of so-called input-nerves. As tests we use strings as follows. Every position in the string corresponds to a position in the vector of input-nerves. A character in the string is either a color or a so-called joker. A teststring is fullled if and only if every character is either equal to the current value of the corresponding input-nerve or a joker. Test-strings are created with three universal operators: Adaption produces a test-string S as \snapshot" of the current environment i.e., the entries of S are set equal to the entries in the input-nerves. Mutation produces a new test-string S from an existing one S 0 by replacing an entry in S 0 by a joker at a random position. Crossover produces a new test-string S from two existing ones S 0 and S 00 by copying the head of S 0 and the tail of S 00 with respect to a random position j. The operators work on a population of tests from SRRs from M sp, the set of SRRs with an edge leading to from the standpoint. In doing so, they focus on SRRs with high tness according to the quality-measures. Gary L. Drescher presents in [2] an own approach to learn eye-hand-coordination. His best run on a Thinking Machines CM2 (16K processors, 512 Mbyte main memory) ended after one day with memory overow. His system learned approximately 70% of the desired world-model. A corresponding world-model is found with Stimulus Response Learning in 25 seconds on a SUN Sparc 10 completely. The total amount of memory used is less than 250 Kbyte. Furthermore, Drescher's experiments are done in simulations only. All experiments with Stimulus Response Learning were done in real world set-ups as well. In doing so, the system was very successful in dealing with noise and errors. 3.4 Unprocessed Real-World Images In the most challenging class of experiments a real-world set-up is used with video images with resolutions up to pixel under changing illumination conditions (weather). Furthermore, no vision processing is used. So, pictures of the hand and the building-blocks are disturbed by reexes, shadows, changing brightness etc. In this class of experiments, tests are represented as programs in a simple turtlegraphics language with commands to move the turtle, to draw lines, and to do a for-loop. A test is fullled if and only if the output of the program is contained in the current video image. Instead of adaption, mutation, and crossover, following operators are used to create turtle-tests: Conc : Takes two programs and concatenates them to one.

6 Split : Splits a program into two new ones at a randomly chosen place. Hill-climbing on constants : Performs a hill-climbing-step on a randomly chosen constant in a program to minimize a picture-distance-function. The picture-distance-function is a heuristic to measure the similarity of two images. Details can be found in [1]. In simulations the system learned very fast hand-movements and grasping. In doing so, it represented the hand as red triangle and building-blocks as rectangles. Due to this promising results we thought adaption to a real world set-up to be easy. We expected usage of vision lters and some fuzziness of tests to be sucient. But a series of experiments showed, that in the real world the camera input contains no triangles or rectangles. The hand and the building-blocks have no simple geometric representation. Noise due to reexes and shadows is always present. An unexpected solution of this problem was found by the system itself. We started a run using unprocessed real world images as input to see what would happen. After few hours the system started to move and grip successfully. It invented a kind of edge-detection. The system represented the hand and building-blocks via programs testing parts of the contour. Following analysis of the camera input revealed that these lines are hardly disturbed in our set-up. Therefore, they can be used by the system to determine the position of the hand and building-blocks. This is realized by the system with the help of the quality-measures. This result was achieved not only once. In every run done so far hand movements and grasping were learned using this edge-detection. The complete world-model for these tasks is learned in approximately 50 hours on average. In doing so, the run-time is dominated by the speed of the robot arm. 4 Conclusion We presented stimulus response learning as a new approach to enable robots to learn an adaptive and robust control-mechanism. This control-mechanism is based on a behavioral model of the world, that is learned with an evolutionary algorithm. The basic algorithm is neither dependent on the robot nor its environment. We presented experiments to learn eye-hand-coordination i.e., to control a robot-arm via a camera. Our results are not from simulations only, but as well from real-world set-ups. References [1] Andreas Birk. Learning geometric concepts with an evolutionary algorithm. In Proc. of The Fifth Annual Conference on Evolutionary Programming. The MIT Press, Cambridge, [2] Gary L. Drescher. Made-up minds, A constructivist approach to articial intelligence. The MIT Press, Cambridge, [3] Jean Piaget. Gesammelte Werke. Klett-Cotta, Stuttgart, 1991.

Learning Geometric Concepts with an Evolutionary Algorithm

Learning Geometric Concepts with an Evolutionary Algorithm final version in: Lawrence J. Fogel, Peter J. Angeline and Thomas Bck (Ed.), Evolutionary Programming V, The MIT Press, 1996 Learning Geometric Concepts with an Evolutionary Algorithm Andreas Birk Universität

More information

An Interactive Technique for Robot Control by Using Image Processing Method

An Interactive Technique for Robot Control by Using Image Processing Method An Interactive Technique for Robot Control by Using Image Processing Method Mr. Raskar D. S 1., Prof. Mrs. Belagali P. P 2 1, E&TC Dept. Dr. JJMCOE., Jaysingpur. Maharashtra., India. 2 Associate Prof.

More information

Generating and Learning from 3D Models of Objects through Interactions. by Kiana Alcala, Kathryn Baldauf, Aylish Wrench

Generating and Learning from 3D Models of Objects through Interactions. by Kiana Alcala, Kathryn Baldauf, Aylish Wrench Generating and Learning from 3D Models of Objects through Interactions by Kiana Alcala, Kathryn Baldauf, Aylish Wrench Abstract For our project, we plan to implement code on the robotic arm that will allow

More information

Genetic Programming of Autonomous Agents. Functional Requirements List and Performance Specifi cations. Scott O'Dell

Genetic Programming of Autonomous Agents. Functional Requirements List and Performance Specifi cations. Scott O'Dell Genetic Programming of Autonomous Agents Functional Requirements List and Performance Specifi cations Scott O'Dell Advisors: Dr. Joel Schipper and Dr. Arnold Patton November 23, 2010 GPAA 1 Project Goals

More information

Genetic Algorithm Based Template Optimization for a Vision System: Obstacle Detection

Genetic Algorithm Based Template Optimization for a Vision System: Obstacle Detection ISTET'09 Umair Ali Khan, Alireza Fasih, Kyandoghere Kyamakya, Jean Chamberlain Chedjou Transportation Informatics Group, Alpen Adria University, Klagenfurt, Austria. Genetic Algorithm Based Template Optimization

More information

Learning Geometric Concepts with an Evolutionary Algorithm. Andreas Birk. Universitat des Saarlandes, c/o Lehrstuhl Prof. W.J.

Learning Geometric Concepts with an Evolutionary Algorithm. Andreas Birk. Universitat des Saarlandes, c/o Lehrstuhl Prof. W.J. Learnin Geometric Concepts with an Evolutionary Alorithm Andreas Birk Universitat des Saarlandes, c/o Lehrstuhl Prof. W.J. Paul Postfach 151150, 66041 Saarbrucken, Germany cyrano@cs.uni-sb.de http://www-wjp.cs.uni-sb.de/cyrano/

More information

Techniques. IDSIA, Istituto Dalle Molle di Studi sull'intelligenza Articiale. Phone: Fax:

Techniques. IDSIA, Istituto Dalle Molle di Studi sull'intelligenza Articiale. Phone: Fax: Incorporating Learning in Motion Planning Techniques Luca Maria Gambardella and Marc Haex IDSIA, Istituto Dalle Molle di Studi sull'intelligenza Articiale Corso Elvezia 36 - CH - 6900 Lugano Phone: +41

More information

Object classes. recall (%)

Object classes. recall (%) Using Genetic Algorithms to Improve the Accuracy of Object Detection Victor Ciesielski and Mengjie Zhang Department of Computer Science, Royal Melbourne Institute of Technology GPO Box 2476V, Melbourne

More information

output vector controller y s... y 2 y 1

output vector controller y s... y 2 y 1 A Neuro-Fuzzy Solution for Fine-Motion Control Based on Vision and Force Sensors Yorck von Collani, Jianwei Zhang, Alois Knoll Technical Computer Science, Faculty of Technology, University of Bielefeld,

More information

Document Image Restoration Using Binary Morphological Filters. Jisheng Liang, Robert M. Haralick. Seattle, Washington Ihsin T.

Document Image Restoration Using Binary Morphological Filters. Jisheng Liang, Robert M. Haralick. Seattle, Washington Ihsin T. Document Image Restoration Using Binary Morphological Filters Jisheng Liang, Robert M. Haralick University of Washington, Department of Electrical Engineering Seattle, Washington 98195 Ihsin T. Phillips

More information

ISSN: [Keswani* et al., 7(1): January, 2018] Impact Factor: 4.116

ISSN: [Keswani* et al., 7(1): January, 2018] Impact Factor: 4.116 IJESRT INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH TECHNOLOGY AUTOMATIC TEST CASE GENERATION FOR PERFORMANCE ENHANCEMENT OF SOFTWARE THROUGH GENETIC ALGORITHM AND RANDOM TESTING Bright Keswani,

More information

Machine Evolution. Machine Evolution. Let s look at. Machine Evolution. Machine Evolution. Machine Evolution. Machine Evolution

Machine Evolution. Machine Evolution. Let s look at. Machine Evolution. Machine Evolution. Machine Evolution. Machine Evolution Let s look at As you will see later in this course, neural networks can learn, that is, adapt to given constraints. For example, NNs can approximate a given function. In biology, such learning corresponds

More information

Camera Parameters Estimation from Hand-labelled Sun Sositions in Image Sequences

Camera Parameters Estimation from Hand-labelled Sun Sositions in Image Sequences Camera Parameters Estimation from Hand-labelled Sun Sositions in Image Sequences Jean-François Lalonde, Srinivasa G. Narasimhan and Alexei A. Efros {jlalonde,srinivas,efros}@cs.cmu.edu CMU-RI-TR-8-32 July

More information

Exploration of Unknown or Partially Known. Prof. Dr. -Ing. G. Farber. the current step. It can be derived from a third camera

Exploration of Unknown or Partially Known. Prof. Dr. -Ing. G. Farber. the current step. It can be derived from a third camera Exploration of Unknown or Partially Known Environments? Darius Burschka, Christof Eberst Institute of Process Control Computers Prof. Dr. -Ing. G. Farber Technische Universitat Munchen 80280 Munich, Germany

More information

joint 3 link 3 link 2 joint 1

joint 3 link 3 link 2 joint 1 Robot Arm Fuzzy Control by a Neuro-Genetic Algorithm Carlos Kavka, Mara Liz Crespo Proyecto UNSL 338403 Departamento de Informatica Universidad Nacional de San Luis Ejercito de los Andes 950, 5700, San

More information

Introducing SB-CoRLA, a Schema-Based Constructivist Robot Learning Architecture

Introducing SB-CoRLA, a Schema-Based Constructivist Robot Learning Architecture Introducing SB-CoRLA, a Schema-Based Constructivist Robot Learning Architecture Yifan Tang and Lynne E. Parker Distributed Intelligence Laboratory, Department of Electrical Engineering and Computer Science

More information

Towards Schema-Based, Constructivist Robot Learning: Validating an Evolutionary Search Algorithm for Schema Chunking

Towards Schema-Based, Constructivist Robot Learning: Validating an Evolutionary Search Algorithm for Schema Chunking Proc. of IEEE International Conference on Robotics and Automation, Pasadena, CA, 2008. Towards Schema-Based, Constructivist Robot Learning: Validating an Evolutionary Search Algorithm for Schema Chunking

More information

An Improved Genetic Algorithm based Fault tolerance Method for distributed wireless sensor networks.

An Improved Genetic Algorithm based Fault tolerance Method for distributed wireless sensor networks. An Improved Genetic Algorithm based Fault tolerance Method for distributed wireless sensor networks. Anagha Nanoti, Prof. R. K. Krishna M.Tech student in Department of Computer Science 1, Department of

More information

VISION-BASED HANDLING WITH A MOBILE ROBOT

VISION-BASED HANDLING WITH A MOBILE ROBOT VISION-BASED HANDLING WITH A MOBILE ROBOT STEFAN BLESSING TU München, Institut für Werkzeugmaschinen und Betriebswissenschaften (iwb), 80290 München, Germany, e-mail: bl@iwb.mw.tu-muenchen.de STEFAN LANSER,

More information

Self-formation, Development and Reproduction of the Artificial System

Self-formation, Development and Reproduction of the Artificial System Solid State Phenomena Vols. 97-98 (4) pp 77-84 (4) Trans Tech Publications, Switzerland Journal doi:.48/www.scientific.net/ssp.97-98.77 Citation (to be inserted by the publisher) Copyright by Trans Tech

More information

Section 1.2: Points and Lines

Section 1.2: Points and Lines Section 1.2: Points and Lines Objective: Graph points and lines using x and y coordinates. Often, to get an idea of the behavior of an equation we will make a picture that represents the solutions to the

More information

Genetic programming. Lecture Genetic Programming. LISP as a GP language. LISP structure. S-expressions

Genetic programming. Lecture Genetic Programming. LISP as a GP language. LISP structure. S-expressions Genetic programming Lecture Genetic Programming CIS 412 Artificial Intelligence Umass, Dartmouth One of the central problems in computer science is how to make computers solve problems without being explicitly

More information

Understanding Tracking and StroMotion of Soccer Ball

Understanding Tracking and StroMotion of Soccer Ball Understanding Tracking and StroMotion of Soccer Ball Nhat H. Nguyen Master Student 205 Witherspoon Hall Charlotte, NC 28223 704 656 2021 rich.uncc@gmail.com ABSTRACT Soccer requires rapid ball movements.

More information

Genetic Programming: A study on Computer Language

Genetic Programming: A study on Computer Language Genetic Programming: A study on Computer Language Nilam Choudhary Prof.(Dr.) Baldev Singh Er. Gaurav Bagaria Abstract- this paper describes genetic programming in more depth, assuming that the reader is

More information

Evolutionary Linkage Creation between Information Sources in P2P Networks

Evolutionary Linkage Creation between Information Sources in P2P Networks Noname manuscript No. (will be inserted by the editor) Evolutionary Linkage Creation between Information Sources in P2P Networks Kei Ohnishi Mario Köppen Kaori Yoshida Received: date / Accepted: date Abstract

More information

Department of Electrical Engineering, Keio University Hiyoshi Kouhoku-ku Yokohama 223, Japan

Department of Electrical Engineering, Keio University Hiyoshi Kouhoku-ku Yokohama 223, Japan Shape Modeling from Multiple View Images Using GAs Satoshi KIRIHARA and Hideo SAITO Department of Electrical Engineering, Keio University 3-14-1 Hiyoshi Kouhoku-ku Yokohama 223, Japan TEL +81-45-563-1141

More information

A Two Stage Real{Time Object Tracking System. Universitat Erlangen{Nurnberg. Martensstr. 3, D{91058 Erlangen, Germany

A Two Stage Real{Time Object Tracking System. Universitat Erlangen{Nurnberg. Martensstr. 3, D{91058 Erlangen, Germany 0 A Two Stage Real{Time Object Tracking System J. Denzler, H. Niemann Lehrstuhl fur Mustererkennung (Informatik 5) Universitat Erlangen{Nurnberg Martensstr. 3, D{91058 Erlangen, Germany email: fdenzler,niemanng@informatik.uni-erlangen.de

More information

Proceedings of the 6th Int. Conf. on Computer Analysis of Images and Patterns. Direct Obstacle Detection and Motion. from Spatio-Temporal Derivatives

Proceedings of the 6th Int. Conf. on Computer Analysis of Images and Patterns. Direct Obstacle Detection and Motion. from Spatio-Temporal Derivatives Proceedings of the 6th Int. Conf. on Computer Analysis of Images and Patterns CAIP'95, pp. 874-879, Prague, Czech Republic, Sep 1995 Direct Obstacle Detection and Motion from Spatio-Temporal Derivatives

More information

Performance Measures for Multi-Graded Relevance

Performance Measures for Multi-Graded Relevance Performance Measures for Multi-Graded Relevance Christian Scheel, Andreas Lommatzsch, and Sahin Albayrak Technische Universität Berlin, DAI-Labor, Germany {christian.scheel,andreas.lommatzsch,sahin.albayrak}@dai-labor.de

More information

Master Thesis: Extraction of Vectorial Information and Accuracy Analysis of Railway Mapping

Master Thesis: Extraction of Vectorial Information and Accuracy Analysis of Railway Mapping Master Thesis: Extraction of Vectorial Information and Accuracy Analysis of Railway Mapping Author: Berdanoglu, Cihat Bertan Supervisor: Prof. Dr. Ing. Dieter Fritsch Co- supervisor: Dr. Ing. Susanne Becker

More information

Vision-based Manipulator Navigation. using Mixtures of RBF Neural Networks. Wolfram Blase, Josef Pauli, and Jorg Bruske

Vision-based Manipulator Navigation. using Mixtures of RBF Neural Networks. Wolfram Blase, Josef Pauli, and Jorg Bruske Vision-based Manipulator Navigation using Mixtures of RBF Neural Networks Wolfram Blase, Josef Pauli, and Jorg Bruske Christian{Albrechts{Universitat zu Kiel, Institut fur Informatik Preusserstrasse 1-9,

More information

A modular neural network architecture for inverse kinematics model learning

A modular neural network architecture for inverse kinematics model learning Neurocomputing 38}40 (2001) 797}805 A modular neural network architecture for inverse kinematics model learning Eimei Oyama*, Arvin Agah, Karl F. MacDorman, Taro Maeda, Susumu Tachi Intelligent System

More information

Prewitt. Gradient. Image. Op. Merging of Small Regions. Curve Approximation. and

Prewitt. Gradient. Image. Op. Merging of Small Regions. Curve Approximation. and A RULE-BASED SYSTEM FOR REGION SEGMENTATION IMPROVEMENT IN STEREOVISION M. Buvry, E. Zagrouba and C. J. Krey ENSEEIHT - IRIT - UA 1399 CNRS Vision par Calculateur A. Bruel 2 Rue Camichel, 31071 Toulouse

More information

Art, Nature, and Patterns Introduction

Art, Nature, and Patterns Introduction Art, Nature, and Patterns Introduction to LOGO Describing patterns with symbols This tutorial is designed to introduce you to some basic LOGO commands as well as two fundamental and powerful principles

More information

1 Lab + Hwk 5: Particle Swarm Optimization

1 Lab + Hwk 5: Particle Swarm Optimization 1 Lab + Hwk 5: Particle Swarm Optimization This laboratory requires the following equipment: C programming tools (gcc, make), already installed in GR B001 Webots simulation software Webots User Guide Webots

More information

Fast Denoising for Moving Object Detection by An Extended Structural Fitness Algorithm

Fast Denoising for Moving Object Detection by An Extended Structural Fitness Algorithm Fast Denoising for Moving Object Detection by An Extended Structural Fitness Algorithm ALBERTO FARO, DANIELA GIORDANO, CONCETTO SPAMPINATO Dipartimento di Ingegneria Informatica e Telecomunicazioni Facoltà

More information

Dept. of Computer Science. The eld of time series analysis and forecasting methods has signicantly changed in the last

Dept. of Computer Science. The eld of time series analysis and forecasting methods has signicantly changed in the last Model Identication and Parameter Estimation of ARMA Models by Means of Evolutionary Algorithms Susanne Rolf Dept. of Statistics University of Dortmund Germany Joachim Sprave y Dept. of Computer Science

More information

Natural Light in Design: IAP 2009

Natural Light in Design: IAP 2009 ECOTECT TUTORIAL --Si Siân Kleindienst-- 1 Getting Started 1.1 -- Entering the License The first time you start Ecotect, it will present you with this screen: You should begin by clicking on activate license

More information

Artificial Intelligence Prof. Deepak Khemani Department of Computer Science and Engineering Indian Institute of Technology, Madras

Artificial Intelligence Prof. Deepak Khemani Department of Computer Science and Engineering Indian Institute of Technology, Madras Artificial Intelligence Prof. Deepak Khemani Department of Computer Science and Engineering Indian Institute of Technology, Madras (Refer Slide Time: 00:17) Lecture No - 10 Hill Climbing So, we were looking

More information

Lighting- and Occlusion-robust View-based Teaching/Playback for Model-free Robot Programming

Lighting- and Occlusion-robust View-based Teaching/Playback for Model-free Robot Programming Lighting- and Occlusion-robust View-based Teaching/Playback for Model-free Robot Programming *Yusuke MAEDA (Yokohama National University) Yoshito SAITO (Ricoh Corp) Background Conventional Teaching/Playback

More information

Programmable Pattern-Formation and Scale-Independence

Programmable Pattern-Formation and Scale-Independence Programmable Pattern-Formation and Scale-Independence Radhika Nagpal PostDoctoral Lecturer, MIT Artificial Intelligence Lab radhi@ai.mit.edu This paper presents a programming language for pattern-formation

More information

Phase2. Phase 1. Video Sequence. Frame Intensities. 1 Bi-ME Bi-ME Bi-ME. Motion Vectors. temporal training. Snake Images. Boundary Smoothing

Phase2. Phase 1. Video Sequence. Frame Intensities. 1 Bi-ME Bi-ME Bi-ME. Motion Vectors. temporal training. Snake Images. Boundary Smoothing CIRCULAR VITERBI BASED ADAPTIVE SYSTEM FOR AUTOMATIC VIDEO OBJECT SEGMENTATION I-Jong Lin, S.Y. Kung ijonglin@ee.princeton.edu Princeton University Abstract - Many future video standards such as MPEG-4

More information

Preprocessing of Stream Data using Attribute Selection based on Survival of the Fittest

Preprocessing of Stream Data using Attribute Selection based on Survival of the Fittest Preprocessing of Stream Data using Attribute Selection based on Survival of the Fittest Bhakti V. Gavali 1, Prof. Vivekanand Reddy 2 1 Department of Computer Science and Engineering, Visvesvaraya Technological

More information

SIMULATION OF ARTIFICIAL SYSTEMS BEHAVIOR IN PARAMETRIC EIGHT-DIMENSIONAL SPACE

SIMULATION OF ARTIFICIAL SYSTEMS BEHAVIOR IN PARAMETRIC EIGHT-DIMENSIONAL SPACE 78 Proceedings of the 4 th International Conference on Informatics and Information Technology SIMULATION OF ARTIFICIAL SYSTEMS BEHAVIOR IN PARAMETRIC EIGHT-DIMENSIONAL SPACE D. Ulbikiene, J. Ulbikas, K.

More information

Using Game Theory for Image Segmentation

Using Game Theory for Image Segmentation Using Game Theory for Image Segmentation Elizabeth Cassell Sumanth Kolar Alex Yakushev 1 Introduction 21st March 2007 The goal of image segmentation, is to distinguish objects from background. Robust segmentation

More information

Evolving Hierarchical and Recursive Teleo-reactive Programs through Genetic Programming

Evolving Hierarchical and Recursive Teleo-reactive Programs through Genetic Programming Evolving Hierarchical and Recursive Teleo-reactive Programs through Genetic Programming Mykel J. Kochenderfer Department of Computer Science Stanford University Stanford, California 94305 mykel@cs.stanford.edu

More information

A Hybrid Approach for Misbehavior Detection in Wireless Ad-Hoc Networks

A Hybrid Approach for Misbehavior Detection in Wireless Ad-Hoc Networks A Hybrid Approach for Misbehavior Detection in Wireless Ad-Hoc Networks S. Balachandran, D. Dasgupta, L. Wang Intelligent Security Systems Research Lab Department of Computer Science The University of

More information

The rest of the paper is organized as follows: we rst shortly describe the \growing neural gas" method which we have proposed earlier [3]. Then the co

The rest of the paper is organized as follows: we rst shortly describe the \growing neural gas method which we have proposed earlier [3]. Then the co In: F. Fogelman and P. Gallinari, editors, ICANN'95: International Conference on Artificial Neural Networks, pages 217-222, Paris, France, 1995. EC2 & Cie. Incremental Learning of Local Linear Mappings

More information

Robot Controller. Sensor Processing Module. Locomotion Module. Sensors. Environment

Robot Controller. Sensor Processing Module. Locomotion Module. Sensors. Environment Evolution of Controllers from a High-Level Simulator to a High DOF Robot G. S. Hornby 1, S. Takamura 2, O. Hanagata 3, M. Fujita 3, and J. Pollack 1 1 Computer Science Dept., Brandeis University, Waltham,

More information

Last update: May 4, Vision. CMSC 421: Chapter 24. CMSC 421: Chapter 24 1

Last update: May 4, Vision. CMSC 421: Chapter 24. CMSC 421: Chapter 24 1 Last update: May 4, 200 Vision CMSC 42: Chapter 24 CMSC 42: Chapter 24 Outline Perception generally Image formation Early vision 2D D Object recognition CMSC 42: Chapter 24 2 Perception generally Stimulus

More information

(a) (b) (c) Fig. 1. Omnidirectional camera: (a) principle; (b) physical construction; (c) captured. of a local vision system is more challenging than

(a) (b) (c) Fig. 1. Omnidirectional camera: (a) principle; (b) physical construction; (c) captured. of a local vision system is more challenging than An Omnidirectional Vision System that finds and tracks color edges and blobs Felix v. Hundelshausen, Sven Behnke, and Raul Rojas Freie Universität Berlin, Institut für Informatik Takustr. 9, 14195 Berlin,

More information

MOTION. Feature Matching/Tracking. Control Signal Generation REFERENCE IMAGE

MOTION. Feature Matching/Tracking. Control Signal Generation REFERENCE IMAGE Head-Eye Coordination: A Closed-Form Solution M. Xie School of Mechanical & Production Engineering Nanyang Technological University, Singapore 639798 Email: mmxie@ntuix.ntu.ac.sg ABSTRACT In this paper,

More information

MOSES - the Mobile Service and Exploration System. Georg Schneider and Martin Greving

MOSES - the Mobile Service and Exploration System. Georg Schneider and Martin Greving MOSES - the Mobile Service and Exploration System Georg Schneider and Martin Greving Abstract. This paper describes the design and implementation of the mobile service and exploration system MOSES, which

More information

Robust-IAC Exploration Explorer for Matlab Release 1.0

Robust-IAC Exploration Explorer for Matlab Release 1.0 Robust-IAC Exploration Explorer for Matlab Release 1.0 Adrien Baranes, Pierre-Yves Oudeyer INRIA Bordeaux - Sud-Ouest 351 Cours de la Liberation, 33405 Talence, France {Adrien.Baranes, Pierre-Yves.Oudeyer}@inria.fr

More information

Optimal tree for Genetic Algorithms in the Traveling Salesman Problem (TSP).

Optimal tree for Genetic Algorithms in the Traveling Salesman Problem (TSP). Optimal tree for Genetic Algorithms in the Traveling Salesman Problem (TSP). Liew Sing liews_ryan@yahoo.com.sg April 1, 2012 Abstract In this paper, the author proposes optimal tree as a gauge for the

More information

A Method of Annotation Extraction from Paper Documents Using Alignment Based on Local Arrangements of Feature Points

A Method of Annotation Extraction from Paper Documents Using Alignment Based on Local Arrangements of Feature Points A Method of Annotation Extraction from Paper Documents Using Alignment Based on Local Arrangements of Feature Points Tomohiro Nakai, Koichi Kise, Masakazu Iwamura Graduate School of Engineering, Osaka

More information

Review: Final Exam CPSC Artificial Intelligence Michael M. Richter

Review: Final Exam CPSC Artificial Intelligence Michael M. Richter Review: Final Exam Model for a Learning Step Learner initially Environm ent Teacher Compare s pe c ia l Information Control Correct Learning criteria Feedback changed Learner after Learning Learning by

More information

Name: Pythagorean Theorem February 3, 2014

Name: Pythagorean Theorem February 3, 2014 1. John leaves school to go home. He walks 6 blocks North and then 8 blocks west. How far is John from the school? 5. A 26 foot long ladder is leaning up against a house with its base 10 feet away from

More information

DeskCNC setup and operation manual

DeskCNC setup and operation manual DeskCNC setup and operation manual This document explains how to install, setup and cut foam shapes using DeskCNC 4 axis foam cutting software. The document will go through a step by step process of how

More information

FLEXIBLE ROBOT-BASED INLINE QUALITY MONITORING USING PICTURE-GIVING SENSORS

FLEXIBLE ROBOT-BASED INLINE QUALITY MONITORING USING PICTURE-GIVING SENSORS FLEXIBLE ROBOT-BASED INLINE QUALITY MONITORING USING PICTURE-GIVING SENSORS Chen-Ko Sung, Andreas Jacubasch and Thomas Müller Fraunhofer Institute IITB, Fraunhoferstrasse 1, 76131 Karlsruhe, Germany chen-ko.sung@iitb.fraunhofer.de,

More information

Mingle Face Detection using Adaptive Thresholding and Hybrid Median Filter

Mingle Face Detection using Adaptive Thresholding and Hybrid Median Filter Mingle Face Detection using Adaptive Thresholding and Hybrid Median Filter Amandeep Kaur Department of Computer Science and Engg Guru Nanak Dev University Amritsar, India-143005 ABSTRACT Face detection

More information

An actor-critic reinforcement learning controller for a 2-DOF ball-balancer

An actor-critic reinforcement learning controller for a 2-DOF ball-balancer An actor-critic reinforcement learning controller for a 2-DOF ball-balancer Andreas Stückl Michael Meyer Sebastian Schelle Projektpraktikum: Computational Neuro Engineering 2 Empty side for praktikums

More information

Towards Cognitive Agents: Embodiment based Object Recognition for Vision-Based Mobile Agents Kazunori Terada 31y, Takayuki Nakamura 31, Hideaki Takeda

Towards Cognitive Agents: Embodiment based Object Recognition for Vision-Based Mobile Agents Kazunori Terada 31y, Takayuki Nakamura 31, Hideaki Takeda Towards Cognitive Agents: Embodiment based Object Recognition for Vision-Based Mobile Agents Kazunori Terada 31y, Takayuki Nakamura 31, Hideaki Takeda 3231 and Tsukasa Ogasawara 31 31 Dept. of Information

More information

Rene Descartes (left) and Pierre de Fermat (right) developed analytical geometry independently of each other in the 1600 s.

Rene Descartes (left) and Pierre de Fermat (right) developed analytical geometry independently of each other in the 1600 s. Lesson 14 Analtical Geometr, Part I Rene Descartes (left) and Pierre de Fermat (right) developed analtical geometr independentl of each other in the 1600 s. Rules and Definitions Rules No new rules for

More information

MIT Programming Contest Team Contest 1 Problems 2008

MIT Programming Contest Team Contest 1 Problems 2008 MIT Programming Contest Team Contest 1 Problems 2008 October 5, 2008 1 Edit distance Given a string, an edit script is a set of instructions to turn it into another string. There are four kinds of instructions

More information

Flexible Gripper System for Small Optical Assemblies Final Tests and Findings

Flexible Gripper System for Small Optical Assemblies Final Tests and Findings Flexible Gripper System for Small Optical Assemblies Final Tests and Findings Timo Prusi 1, Riku Heikkilä 1, T.H. Ha 2, J.Y. Song 2, C.W. Lee 2, Reijo Tuokko 1 1 Tampere University of Technology, Department

More information

EE795: Computer Vision and Intelligent Systems

EE795: Computer Vision and Intelligent Systems EE795: Computer Vision and Intelligent Systems Spring 2012 TTh 17:30-18:45 WRI C225 Lecture 02 130124 http://www.ee.unlv.edu/~b1morris/ecg795/ 2 Outline Basics Image Formation Image Processing 3 Intelligent

More information

Automated Measurement of Viscosity with Ubbelohde Viscometers, Camera Unit and Image Processing Software

Automated Measurement of Viscosity with Ubbelohde Viscometers, Camera Unit and Image Processing Software Santiago de Querétaro, México, Automated Measurement of Viscosity with Ubbelohde Viscometers, Camera Unit and Image Processing Software Will, J.C., Hernández, I., Trujillo, S. km 4,5 Carretera a Los Cués,

More information

Periodic Pattern Detection for Real-Time Application

Periodic Pattern Detection for Real-Time Application Periodic Pattern Detection for Real-Time Application Giovanni Puglisi 1 and Sebastiano Battiato 1 Dipartimento di Matematica e Informatica University of Catania, Italy {puglisi,battiato}@dmi.unict.it Abstract.

More information

ROBOTICS (5 cfu) 09/02/2016. Last and first name Matricola Graduating

ROBOTICS (5 cfu) 09/02/2016. Last and first name Matricola Graduating ROBOTICS (5 cfu) 09/02/2016 Last and first name Matricola Graduating PART 1 - QUIZ (8 marks) 1. For a 3x3 matrix to be a rotation matrix, it should hold - each row vector has module 1 T F - the product

More information

SoMA Product Description

SoMA Product Description SoMA Product Description SoMA Product Description Summary This document is the product description of the Sofica Multimedia Test Automation Solution (SoMA). SoMA is robot aided camera performance test

More information

Understanding Competitive Co-evolutionary Dynamics via Fitness Landscapes

Understanding Competitive Co-evolutionary Dynamics via Fitness Landscapes Understanding Competitive Co-evolutionary Dynamics via Fitness Landscapes Elena Popovici and Kenneth De Jong George Mason University, Fairfax, VA 3 epopovic@gmu.edu, kdejong@gmu.edu Abstract. Co-evolutionary

More information

1 Lab + Hwk 5: Particle Swarm Optimization

1 Lab + Hwk 5: Particle Swarm Optimization 1 Lab + Hwk 5: Particle Swarm Optimization This laboratory requires the following equipment: C programming tools (gcc, make). Webots simulation software. Webots User Guide Webots Reference Manual. The

More information

Two Image-Template Operations for Binary Image Processing. Hongchi Shi. Department of Computer Engineering and Computer Science

Two Image-Template Operations for Binary Image Processing. Hongchi Shi. Department of Computer Engineering and Computer Science Two Image-Template Operations for Binary Image Processing Hongchi Shi Department of Computer Engineering and Computer Science Engineering Building West, Room 331 University of Missouri - Columbia Columbia,

More information

IMPORTANT INSTRUCTIONS

IMPORTANT INSTRUCTIONS 2017 Imaging Science Ph.D. Qualifying Examination June 9, 2017 9:00AM to 12:00PM IMPORTANT INSTRUCTIONS You must complete two (2) of the three (3) questions given for each of the core graduate classes.

More information

CS4758: Rovio Augmented Vision Mapping Project

CS4758: Rovio Augmented Vision Mapping Project CS4758: Rovio Augmented Vision Mapping Project Sam Fladung, James Mwaura Abstract The goal of this project is to use the Rovio to create a 2D map of its environment using a camera and a fixed laser pointer

More information

y 2 x 1 Simulator Controller client (C,Java...) Client Ball position joint 2 link 2 link 1 0 joint 1 link 0 (virtual) State of the ball

y 2 x 1 Simulator Controller client (C,Java...) Client Ball position joint 2 link 2 link 1 0 joint 1 link 0 (virtual) State of the ball Vision-based interaction with virtual worlds for the design of robot controllers D. d'aulignac, V. Callaghan and S. Lucas Department of Computer Science, University of Essex, Colchester CO4 3SQ, UK Abstract

More information

Reinforcement Learning for Appearance Based Visual Servoing in Robotic Manipulation

Reinforcement Learning for Appearance Based Visual Servoing in Robotic Manipulation Reinforcement Learning for Appearance Based Visual Servoing in Robotic Manipulation UMAR KHAN, LIAQUAT ALI KHAN, S. ZAHID HUSSAIN Department of Mechatronics Engineering AIR University E-9, Islamabad PAKISTAN

More information

Complex Systems Made Simple

Complex Systems Made Simple Complex Systems Made Simple 1. Introduction 2. A Complex Systems Sampler 3. Commonalities a. What is NetLogo? b. Graphical interface c. Programming concepts d. Termites NetLogo project? Fall 2015 René

More information

Binocular Tracking Based on Virtual Horopters. ys. Rougeaux, N. Kita, Y. Kuniyoshi, S. Sakane, yf. Chavand

Binocular Tracking Based on Virtual Horopters. ys. Rougeaux, N. Kita, Y. Kuniyoshi, S. Sakane, yf. Chavand Binocular Tracking Based on Virtual Horopters ys. Rougeaux, N. Kita, Y. Kuniyoshi, S. Sakane, yf. Chavand Autonomous Systems Section ylaboratoire de Robotique d'evry Electrotechnical Laboratory Institut

More information

Escaping Local Optima: Genetic Algorithm

Escaping Local Optima: Genetic Algorithm Artificial Intelligence Escaping Local Optima: Genetic Algorithm Dae-Won Kim School of Computer Science & Engineering Chung-Ang University We re trying to escape local optima To achieve this, we have learned

More information

Evolving Turing-Complete Programs for a Register Machine with. Self-modifying Code. 1 Introduction

Evolving Turing-Complete Programs for a Register Machine with. Self-modifying Code. 1 Introduction Evolving Turing-Complete Programs for a Register Machine with Self-modifying Code Peter Nordin Universitat Dortmund Fachbereich Informatik Lehrstuhl fur Systemanalyse D{44221 Dortmund nordin@ls11.informatik.uni-dortmund.de

More information

A Symmetry Operator and Its Application to the RoboCup

A Symmetry Operator and Its Application to the RoboCup A Symmetry Operator and Its Application to the RoboCup Kai Huebner Bremen Institute of Safe Systems, TZI, FB3 Universität Bremen, Postfach 330440, 28334 Bremen, Germany khuebner@tzi.de Abstract. At present,

More information

Rapid Modeling of Digital City Based on Sketchup

Rapid Modeling of Digital City Based on Sketchup Journal of Mechanical Engineering Research and Developments ISSN: 1024-1752 Website: http://www.jmerd.org Vol. 38, No. 1, 2015, pp. 130-134 J. Y. Li *, H. L. Yuan, & C. Reithmeier Department of Architectural

More information

Synchronization and Dissemination in Self-Organizing Communication Networks

Synchronization and Dissemination in Self-Organizing Communication Networks Synchronization and Dissemination in Self-Organizing Communication Networks Christian Bettstetter Networked and Embedded Systems University of Klagenfurt Lakeside Labs GmbH Talk at the Vrije Universiteit

More information

LINE BLOCK SYSTEM APPLICATION NOTE

LINE BLOCK SYSTEM APPLICATION NOTE 1 (19) 16 March 2009 LINE BLOCK SYSTEM APPLICATION NOTE Postaladdress Streetaddress Tel. Fax E-mail Home page P.O. Box 185, FI-00101 Helsinki Kaivokatu 8, 6th floor +358 20 751 5111 +358 20 751 5100 forename.surname@rhk.fi

More information

Hybridization EVOLUTIONARY COMPUTING. Reasons for Hybridization - 1. Naming. Reasons for Hybridization - 3. Reasons for Hybridization - 2

Hybridization EVOLUTIONARY COMPUTING. Reasons for Hybridization - 1. Naming. Reasons for Hybridization - 3. Reasons for Hybridization - 2 Hybridization EVOLUTIONARY COMPUTING Hybrid Evolutionary Algorithms hybridization of an EA with local search techniques (commonly called memetic algorithms) EA+LS=MA constructive heuristics exact methods

More information

Towards a Calibration-Free Robot: The ACT Algorithm for Automatic Online Color Training

Towards a Calibration-Free Robot: The ACT Algorithm for Automatic Online Color Training Towards a Calibration-Free Robot: The ACT Algorithm for Automatic Online Color Training Patrick Heinemann, Frank Sehnke, Felix Streichert, and Andreas Zell Wilhelm-Schickard-Institute, Department of Computer

More information

C. Premsai 1, Prof. A. Kavya 2 School of Computer Science, School of Computer Science Engineering, Engineering VIT Chennai, VIT Chennai

C. Premsai 1, Prof. A. Kavya 2 School of Computer Science, School of Computer Science Engineering, Engineering VIT Chennai, VIT Chennai Traffic Sign Detection Via Graph-Based Ranking and Segmentation Algorithm C. Premsai 1, Prof. A. Kavya 2 School of Computer Science, School of Computer Science Engineering, Engineering VIT Chennai, VIT

More information

A deformable model driven method for handling clothes

A deformable model driven method for handling clothes A deformable model driven method for handling clothes Yasuyo Kita Fuminori Saito Nobuyuki Kita Intelligent Systems Institute, National Institute of Advanced Industrial Science and Technology (AIST) AIST

More information

EXECUTION PLAN OPTIMIZATION TECHNIQUES

EXECUTION PLAN OPTIMIZATION TECHNIQUES EXECUTION PLAN OPTIMIZATION TECHNIQUES Július Štroffek Database Sustaining Engineer Sun Microsystems, Inc. Tomáš Kovařík Faculty of Mathematics and Physics, Charles University, Prague PostgreSQL Conference,

More information

(Refer Slide Time 00:17) Welcome to the course on Digital Image Processing. (Refer Slide Time 00:22)

(Refer Slide Time 00:17) Welcome to the course on Digital Image Processing. (Refer Slide Time 00:22) Digital Image Processing Prof. P. K. Biswas Department of Electronics and Electrical Communications Engineering Indian Institute of Technology, Kharagpur Module Number 01 Lecture Number 02 Application

More information

Sherlock 7 Technical Resource. Search Geometric

Sherlock 7 Technical Resource. Search Geometric Sherlock 7 Technical Resource DALSA Corp., Industrial Products (IPD) www.goipd.com 978.670.2002 (U.S.A.) Document Revision: September 24, 2007 Search Geometric Search utilities A common task in machine

More information

Efficient Grasping from RGBD Images: Learning Using a New Rectangle Representation. Yun Jiang, Stephen Moseson, Ashutosh Saxena Cornell University

Efficient Grasping from RGBD Images: Learning Using a New Rectangle Representation. Yun Jiang, Stephen Moseson, Ashutosh Saxena Cornell University Efficient Grasping from RGBD Images: Learning Using a New Rectangle Representation Yun Jiang, Stephen Moseson, Ashutosh Saxena Cornell University Problem Goal: Figure out a way to pick up the object. Approach

More information

Image Processing algorithm for matching horizons across faults in seismic data

Image Processing algorithm for matching horizons across faults in seismic data Image Processing algorithm for matching horizons across faults in seismic data Melanie Aurnhammer and Klaus Tönnies Computer Vision Group, Otto-von-Guericke University, Postfach 410, 39016 Magdeburg, Germany

More information

Local Search Methods. CS 188: Artificial Intelligence Fall Announcements. Hill Climbing. Hill Climbing Diagram. Today

Local Search Methods. CS 188: Artificial Intelligence Fall Announcements. Hill Climbing. Hill Climbing Diagram. Today CS 188: Artificial Intelligence Fall 2006 Lecture 5: Robot Motion Planning 9/14/2006 Local Search Methods Queue-based algorithms keep fallback options (backtracking) Local search: improve what you have

More information

the change in the context state that occurs (again, prefixed with + or -). As an example: This schema says that when I am in front of a door, and I tr

the change in the context state that occurs (again, prefixed with + or -). As an example: This schema says that when I am in front of a door, and I tr Constructivist Learning: A Neural Implementation of the Schema Mechanism Harold H. Chaput, Benjamin Kuipers and Risto Miikkulainen The University of Texas at Austin 1 University Station C0500 Austin, TX

More information

MLEP Intermediate GPS Workshop Exercise Two Using Maps

MLEP Intermediate GPS Workshop Exercise Two Using Maps During this exercise, you will scale coordinates from a map and enter them into the GPS receiver. This requires a ruler (provided) and all calculations require a paper and pencil. During this exercise,

More information

The NAO Robot, a case of study Robotics Franchi Alessio Mauro

The NAO Robot, a case of study Robotics Franchi Alessio Mauro The NAO Robot, a case of study Robotics 2013-2014 Franchi Alessio Mauro alessiomauro.franchi@polimi.it Who am I? Franchi Alessio Mauro Master Degree in Computer Science Engineer at Politecnico of Milan

More information

A threshold decision of the object image by using the smart tag

A threshold decision of the object image by using the smart tag A threshold decision of the object image by using the smart tag Chang-Jun Im, Jin-Young Kim, Kwan Young Joung, Ho-Gil Lee Sensing & Perception Research Group Korea Institute of Industrial Technology (

More information