Vision-Based Robot Navigation: Quest for Intelligent Approaches Using a Sparse Distributed Memory
Universal-Publishers, 2012 - Technology & Engineering - 240 pages
Starting with a summary of the history of Artificial Intelligence, this book makes the bridge to the modern debate on the definition of Intelligence and the path to building Intelligent Machines. Since the definition of Intelligence is itself subject to open debate, the quest for Intelligent machines is pursuing a moving target. Apparently, intelligent behaviour is, to a great extent, the result of using a sophisticated associative memory, more than the result of heavy processing. The book describes theories on how the brain works, associative memory models and how a particular model - the Sparse Distributed Memory (SDM) - can be used to navigate a robot based on visual memories. Other robot navigation methods are also comprehensively revised and compared to the method proposed. The performance of the SDM-based robot has been tested in different typical problems, such as illumination changes, occlusions and image noise, taking the SDM to the limits. The results are extensively discussed in the book.
What people are saying - Write a review
We haven't found any reviews in the usual places.
A CrossModal Comparison
Acronyms and Symbols
List of Tables
Sparse Distributed Memory
access radius approach architecture arithmetic distance arithmetic mode associative memory autonomous run average bits bitwise mode bitwise operation brain bytes camera captured Chapter co-domain context contrast stretching copies correct data distribution datum described in Section detect encoding environment example experiments follow goal Gray code grey levels Hamming distance hard locations hardware Hierarchical Temporal Memory histogram human image processing implemented input Kanerva kidnapped layer Lizard method MLEs module momentary localisation errors natural binary code navigation algorithm neural network neurons newImg number of hard occupancy grid operation mode optimised code path planning performance pixel possible prediction errors problem proposed requires results obtained retrieve robot navigation robust scenario changes search window sensors sequence number shows similar space Sparse Distributed Memory stored subsumption architecture sum code Summary Table tasks techniques tests topological map vectors view sequences vision-based Voronoi diagram Willshaw Network