Browsing by Author "Jacobs, David"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Development of a Large-Scale Integrated Neurocognitive Architecture Part 1: Conceptual Framework
    (2006-06-15) Reggia, James A.; Tagamets, Malle; Contreras-Vidal, Jose; Weems, Scott; Jacobs, David; Winder, Ransom; Chabuk, Timur
    The idea of creating a general purpose machine intelligence that captures many of the features of human cognition goes back at least to the earliest days of artificial intelligence and neural computation. In spite of more than a half-century of research on this issue, there is currently no existing approach to machine intelligence that comes close to providing a powerful, general-purpose human-level intelligence. However, substantial progress made during recent years in neural computation, high performance computing, neuroscience and cognitive science suggests that a renewed effort to produce a general purpose and adaptive machine intelligence is timely, likely to yield qualitatively more powerful approaches to machine intelligence than those currently existing, and certain to lead to substantial progress in cognitive science, AI and neural computation. In this report, we outline a conceptual framework for the long-term development of a large-scale machine intelligence that is based on the modular organization, dynamics and plasticity of the human brain. Some basic design principles are presented along with a review of some of the relevant existing knowledge about the neurobiological basis of cognition. Three intermediate-scale prototypes for parts of a larger system are successfully implemented, providing support for the effectiveness of several of the principles in our framework. We conclude that a human-competitive neuromorphic system for machine intelligence is a viable long- term goal, but that for the short term, substantial integration with more standard symbolic methods as well as substantial research will be needed to make this goal achievable.
  • Thumbnail Image
    Item
    Segment-based simple-connectivity measure design and implementation
    (2014) Soares, Joao; Baraldi, Andrea; Jacobs, David; Jacobs, David
    In developing different measures for the description of a segment’s shape, we noted that it would be useful to include a measure capable of quantifying the presence of holes. This was motivated by the following scenario. The measures we use to characterize a segment’s shape, such as RoundnessAndNoHole (also known as compactness), ConvexityAndNoHole and RectangularityAndNoHole are monotonically decreasing with the presence of holes, namely: • RoundnessAndNoHole is high if Roundness is high and condition NoHole is true, • ConvexityAndNoHole is high if Convexity is high and condition NoHole is true and, finally, • RectangularityAndNoHole is high if Rectangularity is high and condition NoHole is true. For example, a region with a perfectly round external boundary, but containing several holes, will present a low RoundnessAndNoHole measure. Were the holes not present in the region, it would instead feature a very high RoundnessAndNoHole measure. Besides these measures, our newly introduced version of a measure of elongatedness is also affected by the presence of holes, increasing as the number of holes increases. In our study of satellite images, it is very common to find segments that contain holes, whether due to the underlying holes in the original observed structure or whether due to segmentation errors. In order to reason about these types of situations without having to change the definitions of the shape measures already in use (which are quite natural and intuitive), we introduce a new measure to quantify the presence of holes, which we call simple-connectivity. The simple-connectivity measure quantifies the extent to which a region is simply-connected, i.e., the measure should be monotonically decreasing with holes whose cardinality increases or whose size increases (at fixed cardinality).