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

Undergraduate

Project Type

group

Campus

Daytona Beach

Authors' Class Standing

Poorendra Ramlall, Senior Maximus Langis, Senior Josephine Parker, Senior Raymond Picquet, Senior Leanna Pollicar, Senior Naomi Rodriguez, Senior Simon Sroka, Senior

Lead Presenter's Name

Poorendra Ramlall

Lead Presenter's College

DB College of Arts and Sciences

Faculty Mentor Name

Mihhail Berezovski

Abstract

In order to understand how changes to a material at the atomic and nano-scales impact the way a material behaves, it is necessary to measure the material at that scale. Scanning transmission Electron Microscopy (STEM) is one such approach. Over the last 5 years, computer vision has been completely revolutionized by the advent of deep neural networks (DNN). One challenge of using DNNs for image processing is that they require often large amounts of labelled data in order to build a usable algorithm (model). Researchers at Pacific Northwest National Laboratory (PNNL) have developed a few-shot algorithm which allows the model to be operated using limited data. Using data from a sample micrograph, a new neural network is implemented with the objective of aptly segmenting the micrograph and obtaining classifications for the microstructures. This model investigates the use of existing image segmentation techniques, particularly region-based techniques. The success of this approach can provide a rapid and reconfigurable tool for identifying these microstructures. This research conducted in collaboration with Pacific Northwest National Laboratory.

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
 

Characterization of Material Micro & Nano Structures Using Machine Learning Algorithms

In order to understand how changes to a material at the atomic and nano-scales impact the way a material behaves, it is necessary to measure the material at that scale. Scanning transmission Electron Microscopy (STEM) is one such approach. Over the last 5 years, computer vision has been completely revolutionized by the advent of deep neural networks (DNN). One challenge of using DNNs for image processing is that they require often large amounts of labelled data in order to build a usable algorithm (model). Researchers at Pacific Northwest National Laboratory (PNNL) have developed a few-shot algorithm which allows the model to be operated using limited data. Using data from a sample micrograph, a new neural network is implemented with the objective of aptly segmenting the micrograph and obtaining classifications for the microstructures. This model investigates the use of existing image segmentation techniques, particularly region-based techniques. The success of this approach can provide a rapid and reconfigurable tool for identifying these microstructures. This research conducted in collaboration with Pacific Northwest National Laboratory.

 

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.