You are here

Semantic Models for the Identification of Laboratory Equipment (SMILE)

Award Information
Agency: Department of Defense
Branch: Defense Threat Reduction Agency
Contract: HDTRA120P0003
Agency Tracking Number: T19B-002-0003
Amount: $149,974.00
Phase: Phase I
Program: STTR
Solicitation Topic Code: DTRA19B-002
Solicitation Number: 19.B
Solicitation Year: 2019
Award Year: 2020
Award Start Date (Proposal Award Date): 2019-12-30
Award End Date (Contract End Date): 2020-07-31
Small Business Information
625 Mount Auburn Street
Cambridge, MA 02138
United States
DUNS: 115243701
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Steve Hookway
 Senior Software Engineer
 (617) 491-3474
Business Contact
 Yvonne Fuller
Phone: (617) 491-3474
Research Institution
 University of Wisconsin-Madison
 Brenda A. Egan Brenda A. Egan
21 N. Park St STE 6401
Madison, WI 53715
United States

 (608) 890-3301
 Nonprofit College or University

Military operators must identify and catalogue the equipment they find when inspecting laboratory facilities. This information is used to determine the lab’s capabilities, including the lab’s potential for building weapons of mass destruction. Currently, operators use computer vision algorithms to help them classify equipment in pictures of laboratory environments. Unfortunately, current image processing algorithms have several shortcomings that prevent them from achieving accurate results in this domain. Charles River Analytics and the University of Wisconsin-Madison propose to design and develop Semantic Models for the Identification of Laboratory Equipment (SMILE). The SMILE framework combines state of the art Convolutional Neural Network (CNN) object detection with a semantic model of laboratory capability to detect and classify the laboratory equipment in an image and to provide an estimate of the laboratory’s capability.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government