Adam Byrne
Adam Byrne

I’m Adam, a [ISE] student focused on Computer Vision with Industrial automation and Automotive perception experience on high-level projects. Now working on High Powered Rocketry at AeroSoc UL for the MACH24 competition.

Python logo
C/C++ logo
PyTorch logo
TensorRT logo
ONNX logo

Projects

More
Modular monoliths architected sensor aggregator over MQTT for deployable UAV CanSat payload with multiple sensors and peripherals for Mach24 competition.
cGAN ML model using RGB images to generate NIR images for NDVI and NDWI calculations to evaluate crop health.
Adapted sliding tiled inference for small object detection with Radar using the ONNX runtime. This project was completed as part of my residency at Provizio.
This project received a bronze medal on Kaggle. I developed a classifier to estimate the risk of heart attack per patient based on their medical history.

Posts

View All
3 min read
5D Perception™️ - Radar only classification, tracking and freespace segmentation for autonomous vehicles in all weather conditions.
2 min read
Enabling real-time remote control of UR6e robotic arm using your hand, for increased efficiency on assembly lines without shutdowns.

Career

CV
ULAS HiPR

Dec 2023 - Present

Developing a real time processing chain for a CanSat along with a cGAN ML model to do agri evaluation with an RGB camera. I work alongside the Electronics and Software team leads to create an intelligent CanSat.
Provizio

Sep - Dec 2023

Working with 5D Perception™️ to reduce road deaths to zero. Integrated SAHI ONNX runtime, Parallelized k-fold cross validation and developed profiling tools for the AI team.
Analog Devices

Jun - Aug 2023

Worked towards Industry 4.0 by developing a teleoperation solution for robotic arms using realtime wrist tracking via LiDAR and Infrared to reduce downtime.
ToDesktop

Sep 2021 - Jul 2022

Created a cloud storage optimisation tool to cross reference active customers with files to reduce unused storage. I also worked on the Desktop UI for configuring the look and feel of apps.