You are here

5 – An Approach to Rapidly Curate Large Image Datasets to Train Machine Learning Ship Classification Models

Award Information
Agency: Department of Defense
Branch: Navy
Contract: N68335-20-F-0123
Agency Tracking Number: N193-A02-0341
Amount: $149,961.00
Phase: Phase I
Program: SBIR
Solicitation Topic Code: N193-A02
Solicitation Number: 19.3
Timeline
Solicitation Year: 2019
Award Year: 2020
Award Start Date (Proposal Award Date): 2019-11-21
Award End Date (Contract End Date): 2020-04-20
Small Business Information
185 South Broad St. Suite 303 P.O. Box 1060
Pawcatuck, CT 06379
United States
DUNS: 125370176
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Thomas Santos
 Chief Technology Officer
 (401) 847-3399
 tsantos@rite-solutions.com
Business Contact
 Chad Seelig
Phone: (401) 847-3399
Email: cseelig@rite-solutions.com
Research Institution
N/A
Abstract

A critical bottleneck in machine learning efforts continues to be either the lack of sufficiently sized, fully curated data-sets, or availability of the time and resources required to develop the required data-sets/models via manual identification and tagging. Because large curated data-sets are essential to ship identification and classification using machine learning, Rite-Solutions proposes an approach using weakly supervised learning to automatically generate labels for non-curated data-sets to train ship recognition and classification ML models.Several tools exist on the market and in academia that have a range of capabilities that, when integrated, will provide Weak Supervision, e.g., an automated way of curating data-sets. Weakly supervised machine learning shows strong potential to accurately perform the label and training functions through automation and thereby reduce SME effort and time to develop the models required for high confidence threat/non-threat vessel identification. Equally important, this approach will rapidly incorporate new images and data to aid the warfighter in identifying and classifying new and changing threats.

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

US Flag An Official Website of the United States Government