Search

Your search keyword '"Johnson, Christi"' showing total 3 results
Start Over Author "Johnson, Christi" Search Limiters Full Text
3 results on '"Johnson, Christi"'

1. Expanding Accurate Person Recognition to New Altitudes and Ranges: The BRIAR Dataset

Author

Cornett III, David, Brogan, Joel, Barber, Nell, Aykac, Deniz, Baird, Seth, Burchfield, Nick, Dukes, Carl, Duncan, Andrew, Ferrell, Regina, Goddard, Jim, Jager, Gavin, Larson, Matt, Murphy, Bart, Johnson, Christi, Shelley, Ian, Srinivas, Nisha, Stockwell, Brandon, Thompson, Leanne, Yohe, Matt, Zhang, Robert, Dolvin, Scott, Santos-Villalobos, Hector J., and Bolme, David S.

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Artificial Intelligence, Computer Science - Machine Learning
Abstract

Face recognition technology has advanced significantly in recent years due largely to the availability of large and increasingly complex training datasets for use in deep learning models. These datasets, however, typically comprise images scraped from news sites or social media platforms and, therefore, have limited utility in more advanced security, forensics, and military applications. These applications require lower resolution, longer ranges, and elevated viewpoints. To meet these critical needs, we collected and curated the first and second subsets of a large multi-modal biometric dataset designed for use in the research and development (R&D) of biometric recognition technologies under extremely challenging conditions. Thus far, the dataset includes more than 350,000 still images and over 1,300 hours of video footage of approximately 1,000 subjects. To collect this data, we used Nikon DSLR cameras, a variety of commercial surveillance cameras, specialized long-rage R&D cameras, and Group 1 and Group 2 UAV platforms. The goal is to support the development of algorithms capable of accurately recognizing people at ranges up to 1,000 m and from high angles of elevation. These advances will include improvements to the state of the art in face recognition and will support new research in the area of whole-body recognition using methods based on gait and anthropometry. This paper describes methods used to collect and curate the dataset, and the dataset's characteristics at the current stage.

Published
2022

2. The Mertens Unrolled Network (MU-Net): A High Dynamic Range Fusion Neural Network for Through the Windshield Driver Recognition

Author

Ruby, Max, Bolme, David S., Brogan, Joel, Cornett III, David, Delgado, Baldemar, Jager, Gavin, Johnson, Christi, Martinez-Mendoza, Jose, Santos-Villalobos, Hector, and Srinivas, Nisha

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning, Electrical Engineering and Systems Science - Image and Video Processing
Abstract

Face recognition of vehicle occupants through windshields in unconstrained environments poses a number of unique challenges ranging from glare, poor illumination, driver pose and motion blur. In this paper, we further develop the hardware and software components of a custom vehicle imaging system to better overcome these challenges. After the build out of a physical prototype system that performs High Dynamic Range (HDR) imaging, we collect a small dataset of through-windshield image captures of known drivers. We then re-formulate the classical Mertens-Kautz-Van Reeth HDR fusion algorithm as a pre-initialized neural network, which we name the Mertens Unrolled Network (MU-Net), for the purpose of fine-tuning the HDR output of through-windshield images. Reconstructed faces from this novel HDR method are then evaluated and compared against other traditional and experimental HDR methods in a pre-trained state-of-the-art (SOTA) facial recognition pipeline, verifying the efficacy of our approach., Comment: Accepted to SPEI Autonomous Systems: Sensors, Processing and Security for Vehicles & Infrastructure 2020

Published
2020

3. Recruitment to a Randomized Web-Based Nutritional Intervention Trial: Characteristics of Participants Compared to Non-Participants

Author

Stopponi, Melanie A, Alexander, Gwen L, McClure, Jennifer B, Carroll, Nikki M, Divine, George W, Calvi, Josephine H, Rolnick, Sharon J, Strecher, Victor J, Johnson, Christine Cole, and Ritzwoller, Debra P

Subjects
Computer applications to medicine. Medical informatics, R858-859.7, Public aspects of medicine, RA1-1270
Abstract

Background Web-based behavioral programs efficiently disseminate health information to a broad population, and online tailoring may increase their effectiveness. While the number of Internet-based behavioral interventions has grown in the last several years, additional information is needed to understand the characteristics of subjects who enroll in these interventions, relative to those subjects who are invited to enroll. Objective The aim of the study was to compare the characteristics of participants who enrolled in an online dietary intervention trial (MENU) with those who were invited but chose not to participate, in order to better understand how these groups differ. Methods The MENU trial was conducted among five health plans participating in the HMO Cancer Research Network in collaboration with the University of Michigan Center for Health Communication Research. Approximately 6000 health plan members per site, between the ages of 21 and 65, and stratified by gender with oversampling of minority populations, were randomly selected for recruitment and were mailed an invitation letter containing website information and a US$2 bill with the promise of US$20 for completing follow-up surveys. Administrative and area-based data using geocoding along with baseline survey data were used to compare invitees (HMO members sent the introductory letter), responders (those who entered a study ID on the website), and enrollees (those who completed the enrollment process). Generalized estimating equation multivariate and logistic regression models were used to assess predictors of response and enrollment. Results Of 28,460 members invited to participate, 4270 (15.0%) accessed the website. Of the eligible responders, 2540 (8.9%) completed the consent form and baseline survey and were enrolled and randomized. The odds of responding were 10% lower for every decade of increased age (P < .001), while the likelihood of enrolling was 10% higher for every decade increase in age (P < .001). Women were more likely to respond and to enroll (P < .001). Those living in a census tract associated with higher education levels were more likely to respond and enroll, as well as those residing in tracts with higher income (P < .001). With a 22% (n = 566) enrollment rate for African Americans and 8% (n = 192) for Hispanics, the enrolled sample was more racially and ethnically diverse than the background sampling frame. Conclusions Relative to members invited to participate in the Internet-based intervention, those who enrolled were more likely to be older and live in census tracts associated with higher socioeconomic status. While oversampling of minority health plan members generated an enrolled sample that was more racially and ethnically diverse than the overall health plan population, additional research is needed to better understand methods that will expand the penetration of Internet interventions into more socioeconomically diverse populations. Trial Registration Clinicaltrials.gov NCT00169312; http://clinicaltrials.gov/ct2/show/NCT00169312 (Archived by WebCite at http://www.webcitation.org/5jB50xSfU)

Published
2009
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results