MindsEye Software Technology & Applications

DSS Lab has been researching, implementing and patenting technology for intelligent applications under its MindsEye Software division.

MindsEye Software's recognition technology comes from decades of work across multiple levels of abstraction:

Software products
Software languages
Data models
Foundations of logic and mathematics

MindsEye Software's data model and its underlying typing system provide a unique foundation for solving problems of representation, recognition and learning.
Patent claims have been filed for the following:

Multi-modal recognition
Segment extraction from images
Object, surface and segment extraction from tactile sensations
Spreadsheet interpretation
Term extraction from NL text

MindsEye Software's recognition system provides a method for storing multi-level multi-modal structures in a normalized fashion that, in conjunction with our recognition algorithms, provides for fast recognition of arbitrary object, events or patterns based on partial information. The library manages so-called language objects, such as spreadsheet references, in the same fashion as so-called physical objects such as image patterns. Unique attributes of MindsEye Software's recognition system:

Represents so-called "structures" and "processes", as well as "words" and "things" within the same definitional framework
Represents "recognitional" patterns and "performance" patterns in the same definitional framework
Provides for multilevel multi-modal learning and recognition
Parsing a language scene is equivalent to parsing a visual scene
Spans the gap between bottom-up robotics and top-down symbolic processors

MindsEye Software is interested in partnering for commercial delivery of applications based on the recognition system. Interested parties should contact us for further information.

Applications:


Voyager spreadsheet interpreter
Introduction
Architecture overview

TrueVue image recognizer
Interpretation engine overview