EXPO21XX Online Trade Fair  Navigation : EXPO21XX.com > AUTOMATION 21XX > Universities - Research in Robotics > The George Washington University
Home Promote your business About usContact us
Home back to hall map back to hall   
Go to Search Go to Search

A8    USA The George Washington University
Show Screen
Exit to Hall
Company Info
Visit our Website
  • Machine Intelligence and Cognition
  • Algorithms and Theory
  • Bioinformatics and Biomedical Computing
  • Computer Security and Information Assurance
  • Digital Media
  • Networking and mobile computing
  • Pervasive Computing and Embedded Systems
  • Software Engineering and Systems
Research Overview
ALISA Introduction
Click here to see a new video: Peter Bock discusses the evolution of Artificial Intelligence and states his intention of creating the world's first artificial being. He shows the capabilities of the current system, known as ALISA, in creating works of art.

TEDxGWU - Peter Bock - Emergence of Creativity in Artificial Intelligence

The statistical-learning paradigm known as Collective Learning Systems (CLS) Theory, initially proposed and published in the 1970s by Peter Bock at The George Washington University (GWU) in Washington DC, made its practical engineering debut in the early 1990s as an image classification tool for detecting defective components in industrial automation processes. Known as ALISA (Adaptive Learning Image and Signal Analysis), the first module (the Texture Module) of this classification engine was developed by Professor Bock and his team of doctoral students at the Research Institute for Applied Knowledge Processing (FAW) in Ulm, Germany. Since then ALISA has been continuously refined and extended at GWU, currently enabling it to process images at progressively higher and higher symbolic levels from textures and colors, to geometries, shapes, and components. In addition, ALISA now also includes a similar capability to process a corresponding hierarchy of properties in multi-channel data streams (signals).

The current ALISA system provides five different kinds of processing modules: the Texture Module, the Geometry Module, the Vector Module, the Component Module, and the Shape Module. ALISA system configurations for most practical applications can now run on a typical desktop computer in real-time, e.g., processing standard video at normal frame rates.

Major funding for this 20-year R&D project was provided by the German industrial firm Robert Bosch GmbH throughout the 1990s, resulting in many of the improvements and enhancements of the ALISA system. Since 1990 Project ALISA has been continuously funded for specific practical applications by several US industrial firms and government agencies, including most recently by the Defense Threat Reduction Agency (DoD) for the identification of radiological weapons and the high-speed detection of handguns in x-rays of baggage, despite intentional shielding, partial obscuration, and/or disassembly.

CLS Theory and the design of the ALISA system were biologically motivated by the multi-path architecture and adaptive functions of the visual cortex and frontal lobes in the primate brain. As a natural extension to the current ALISA capabilities, a Lexical Module is now under development to provide intelligent language-independent search of textual information. Long-term research focuses upon the extension of ALISA to human-like levels of symbolic and self-referential cognition, using feedback from the higher to lower levels for disambiguation, as well as limbic response for the evaluation of perceived contextual significance and risk (emotion). Some initial experiments have applied ALISA to the creation of art (see examples at the end below).
ALISA Texture Module
The video shows motion tracking by ALISA.
ALISA Segmentation Module
ALISA Vector Module
ALISA Geometry Module (Canonical)
ALISA Geometry Module (Secular)
ALISA Component module
ALISA Shape Module
ALISA Action Module
The video shows an example of how ALISA paints.

Exit the Hall  Exit to Hall The George Washington University   Exit to Hall   Next Corridor 6-10