Matt Brown
Principal Machine Learning Scientist at Rocket Lab
Cary, North Carolina, United States
Summary
👤
Senior🎓
Top SchoolMatt Brown is a Principal Machine Learning Scientist with nine years of experience building and leading teams that deploy state-of-the-art computer vision and AI for mission-critical space and terrestrial systems. He combines deep expertise in the physics of imaging with practical experience integrating hardware and real-time ML on constrained compute (GPUs and Xilinx Versal), enabling low-SNR detection and tracking that approach theoretical performance bounds. A proven leader and founder, he has run multidisciplinary programs and managed a multi-million-dollar IRAD portfolio tied to aggressive revenue roadmaps while delivering research as principal investigator on government programs (DARPA, IARPA, NVESD). At Kitware and GEOST he designed distributed surveillance and autonomous payload systems that fuse multi-modal deep networks for detection, tracking, and re-identification, and he contributes to open-source photogrammetry tooling such as TeleSculptor. Based in Cary, NC, he pairs a Princeton PhD in Mechanical Engineering with hands-on engineering roots in aerospace optics, bringing uncommon fluency across sensors, algorithms, and deployable systems.
9 years of coding experience
14 years of employment as a software developer
B.S., Mechanical Engineering (minor Mathematics), B.S., Mechanical Engineering (minor Mathematics) at Rutgers, The State University of New Jersey-New BrunswickPh.D., Mechanical Engineering, Ph.D., Mechanical Engineering at Princeton University
Github Skills (9)
csv-parsing10
pandas10
csv-parse10
python10
photogrammetry10
imagery9
cpp4
gui-application4
cplus4
Programming languages (6)
Github contributions (5)
Kitware/TeleSculptor
Mar 2017 - Mar 2017
TeleSculptor: Aerial Photogrammetry Application powered by KWIVER
Role in this project:
Back-end Developer
Back-end DeveloperContributions:5 commits, 1 comment, 2 issues in 13 days
Contributions summary:Matt contributed to the project by implementing and refining a script that converts Sussex MUTC metadata into POS metadata. This involved parsing data from various CSV files (RawImageTopic, pva, and pos_1) and extracting relevant information for each image frame. The user added command-line parsing capabilities, enhanced the script's functionality, and fixed copyright-related issues to improve the project's usability and maintainability.
kitwarephotogrammetryc-plus-plusimage-processingaerial-imagery
Kitware/keypointgui
Sep 2017 - Sep 2021
Provides a wxPython GUI to select image-point correspondences between images and fits a homography.
Contributions:52 commits, 5 PRs, 6 pushes in 4 years
pythonhomographywxpythonguipoint
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.