eCortex

eCortex Awarded $1,500,000 Subcontract By HRL Laboratories, LLC

dave — Wed, 06 Apr 2011 13:31:56 +0000

BOULDER, CO – March 8, 2011 – eCortex, Inc., exclusive commercial licensee of the Emergent(tm) neural network simulation system, was awarded a $1,500,000, 3.5 year subcontract to develop and support Emergent for HRL Laboratories, LLC in its ICArUS (Integrated Cognitive-Neuroscience Architectures for Understanding Sensemaking) program. The subcontract includes a one year initial period of performance with options for subsequent periods and phases.

Supported by the intelligence Advanced Research Projects Activity (IARPA) via Department of the
Interior (DOI) contract number D10PC20021. The U.S Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright annotation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of IARPA, DOI, or the U.S. Government.

eCortex Awarded $347,000 IARPA Research Contract

admin — Wed, 15 Dec 2010 04:04:08 +0000

2010, eCortex Awarded $347,000 for IARPA Research Contract. April 30, 2010 CU licensee and POCi recipient eCortex, Inc. was awarded a one-year contract to use its Leabra/Emergent-based brain models to explore the effect of embodying a learning model in a three-dimensional virtual environment. The title of the project is “Embodied Common Sense in Vision and Language,” and it is sponsored by IARPA (Intelligence Advanced Research Projects Agency) and the U.S. Army Research Office. This summary does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred.

Overview

admin — Wed, 15 Dec 2010 01:34:53 +0000

eCortex is a research and development company dedicated to the idea that the most important advances in artificial intelligence will come from a deep understanding of the mechanisms in the human brain.

We perform basic and applied research for defense and intelligence agencies in programs where the challenges are broad enough to defy narrow algorithmic solutions. Ultimately, we aim to develop commercial products for robotics and analysis of image, video, and other data.

The company was founded in 2006. It is the exclusive commercial licensee of the Emergent neural network simulator and IDE, along with models of the human visual system originally developed at the University of Colorado, Boulder.

Synaptic weights adapt to recognize objects

admin — Mon, 22 Nov 2010 23:52:20 +0000

The network shown is a model of vision with an architecture similar to the organization of the brain’s ventral visual pathway, used to recognize visual objects. Like most neural models, it learns about its subject matter by modifying the strength the connection between units (the smallest boxes in the figure).

Network learns names and semantic information

lee — Mon, 22 Nov 2010 23:51:36 +0000

The object recognition pathway is modeled as a hierarchy of layers that are increasingly invariant to location, rotation, and size. At the highest level, individual units represent object categories and are assigned names; these units also connect to a network of semantic relationships among different categories.