Is this project an undergraduate, graduate, or faculty project?

Undergraduate

Project Type

group

Campus

Daytona Beach

Authors' Class Standing

Ying Zheng, Sophomore Lilian Borchardt, Junior Theryn Compoc Senior Brian Danaher, Senior

Lead Presenter's Name

Brian Danaher

Faculty Mentor Name

Mihhail Berezovski

Abstract

The Signal Processing and Applied Mathematics research group at the Nevada National Security Site (NNSS) is pleased to partner with students at Embry-Riddle Aeronautical University (ERAU) to study and develop a novel pooling method for convolutional neural networks (CNN), using a publicly available data set on diabetic retinopathy images. While most image data sets perform well under existing CNN architectures, there are still challenging datasets which require further advancements, prompting us to develop new techniques for CNNs in order to satisfactorily train a network. NNSS have developed a new pooling method called "variable stride" that demonstrates advantages for networks to train faster and more efficiently. The goal of this partnership is to provide feedback and evaluation of new techniques and studying the effectiveness of this method, compared to other pooling methods.

Did this research project receive funding support (Spark, SURF, Research Abroad, Student Internal Grants, Collaborative, Climbing, or Ignite Grants) from the Office of Undergraduate Research?

No

Share

COinS
 

Enhanced Deep Learning with Applications to Diabetic Retinopathy and National Security

The Signal Processing and Applied Mathematics research group at the Nevada National Security Site (NNSS) is pleased to partner with students at Embry-Riddle Aeronautical University (ERAU) to study and develop a novel pooling method for convolutional neural networks (CNN), using a publicly available data set on diabetic retinopathy images. While most image data sets perform well under existing CNN architectures, there are still challenging datasets which require further advancements, prompting us to develop new techniques for CNNs in order to satisfactorily train a network. NNSS have developed a new pooling method called "variable stride" that demonstrates advantages for networks to train faster and more efficiently. The goal of this partnership is to provide feedback and evaluation of new techniques and studying the effectiveness of this method, compared to other pooling methods.

 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.