Edward Scott Email & Phone Number

Woburn, Massachusetts, United States

Boston, Ma

Teaching assistant for Parallel Processing for Data Analytics.* Teach fundamentals of Apache Spark and its application to data analytics and machine learning on a cluster.* Conduct tutorials, provide instruction during office hours, and grade assignments for 50 students.

* Research human perception of image usefulness, and transfer knowledge to algorithm development.* Apply utility estimation to photographic systems, e.g. to reduce image transmission or storage requirements.* Developed Multi-Scale Difference of Gaussian Utility (MS-DGU), a state-of-the-art full-reference utility estimation algorithm applying keypoint matching in difference-of-gaussian scale space.* Optimized JPEG quantization tables for rate-utility with MS-DGU as the utility measure, achieving a 20% reduction in image size with no loss in predicted utility and no increase in algorithmic complexity.* Developed a convolutional neural network for no-reference utility estimation, implemented with the Caffe framework, with comparable performance to MS-DGU but without requiring an undistorted reference image.

* Developed target detection algorithms for airborne intelligence, surveillance, and reconnaissance radars. * Provided software tools and analysis to monitor radar status and functionality on-site during field tests.* Applied signal processing skills to evaluate radar system performance on a 140TB dataset.* Held Top Secret security clearance.

Evanston, Il

* Investigated machine perception of images, applying machine learning and signal processing techniques.* Designed and ran human subject experiments to objectively evaluate performance of software tools.

Advisors: Thrasos Pappas, Bryan PardoThesis: Interactive Evaluation of Parameters for Photographic Adjustment and Texture* Designed and developed an adaptive user interface capable of learning user-specific transformations from high-level, arbitrary image concepts to low-level image parameters.* Applied interface to photographic editing, automatically generating adjustment sliders customized to user-specific concepts after a brief training which requires no prior editing knowledge.* Explored perceived roughness of stochastic visual textures, relating coefficients of textures’ steerable pyramid decomposition to roughness.* Designed and ran human subject experiments to objectively evaluate performance of software tools.

Phd Candidate (Not Completed), Electrical Engineering

Research Focus: Automatic recognition of image usefulness in task-based imaging scenarios, such as viewing of surveillance imagery..

Ms, Electrical Engineering And Computer Science

Bs, Electrical Engineering