Ben Eakins Email & Phone Number

Melbourne, Fl, Us

Developed a Natural Language Processing algorithm to identify the relationships between entities (people, governments, militant organizations, etc.) described by sentences taken from US Department of State publications, news articles, and Wikipedia.Wrote a novel Neural Network architecture for predicting types of organisms based on protein sequences sampled from the organism.Developed a prototype for a multi-sensor audio detection, characterization, and localization system. Tested the system in an end-to-end simulation.Trained a Generative AI model to perform image compression. The model was constructed so that the compression Neural Network can fit on an edge device such as a Coral AI Dev Board.

Trained an Isolation Forest anomaly detector to flag atypical and poor-quality GPS tracks.Developed a Random Forest method for identifying a specific type of mover using GPS data alone.Wrote a custom CNN (Convolutional Neural Network) architecture in Tensorflow/Keras for performing binary classification on image chips. The network significantly out-performed Resnet50 on the data set.Developed a CNN detector for Radio Frequency (RF) signals in wide-band collections that simultaneously estimated center frequencies and bandwidths.Trained a Unit Box model to detect and classify objects-of-interest in low quality images.Wrote a white paper describing a novel application of Deep Neural Networks for RF transceivers.Developed a novel blind filtering algorithm for separating interfering RF signals based on their location. This work resulted in a patent.

Doctor'S Degree, Physics

Bachelor Of Science, Mathematics

Associate Of Arts - Aa, Physics

Activities and Societies: Phi Theta Kappa