Josh Veitch-michaelis
Postdoctoral Researcher at University of Florida
Gainesville, Florida, Antarctica
Summary
👤
Senior🎓
Top SchoolJosh Veitch-michaelis is a postdoctoral researcher and 12-year imaging and computer vision specialist who blends academic rigor—PhD in Stereo Imaging/LIDAR data fusion from UCL and a 1st class MPhys from Warwick—with hands-on systems engineering in C++ and Python. He has built and deployed embedded and cloud-enabled 3D vision systems for drones, industry, and remote sensing applications, and has wintered twice supporting science operations at polar observatories. His work spans stereo cameras, LiDAR fusion, electronics and data science, and he contributes to high-profile open-source ML tooling (including improvements to the Darknet/YOLO training pipeline). Based in Gainesville but comfortable operating in extreme environments, he focuses on ML for social good, applying advanced imaging and robustness improvements to ecological monitoring and disaster response.
12 years of coding experience
7 years of employment as a software developer
Masters, Physics, First Class, Masters, Physics, First Class at University of WarwickDoctor of Philosophy (PhD), Imaging Science, Doctor of Philosophy (PhD), Imaging Science at University College London, U. of London
Github Skills (7)
c1710
object-detection10
computer-vision10
deep-learning10
c1110
neural-network8
deep-neural-networks8
Programming languages (16)
Github contributions (5)
AlexeyAB/darknet
Mar 2019 - Mar 2019
YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:7 commits, 3 PRs, 7 comments in 4 days
Contributions summary:Josh contributed to the core functionality of the darknet project by modifying code related to annotation file handling, specifically for image formats like TIFF. Their work included adding support for the SGDR learning rate policy, which is crucial for training deep learning models. Additionally, the user made minor code improvements to improve robustness and error messages. The majority of the changes focus on improving the deep learning training process within the repository.
linuxneural-networkobject-detectionwindowsyolov4
i3drobotics/stereo-vision-toolkit
Jul 2017 - Sep 2020
Stereo image processing suite
Contributions:2 releases, 64 commits, 43 pushes in 3 years 3 months
image-processingstereo-visioncamera-calibrationstereocamera
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.