Lakshantha Dissanayake
Embedded Computer Vision Engineer at Ultralytics
Vancouver, British Columbia, Canada
Summary
🤩
Rockstar🎓
Top SchoolLakshantha Dissanayake is an Embedded Computer Vision Engineer with 8 years of hands-on experience deploying and optimizing vision models on edge platforms like NVIDIA Jetson, Raspberry Pi, and Edge TPU. Currently at Ultralytics, he focuses on exporting and hardware-accelerating YOLO models—fixing ARM64/Jetson build issues and adding device-aware export options that improve real-world performance. Previously an application engineer at Seeed Studio and a Udemy instructor for 12K+ learners, he blends product-focused demos, technical support, and educational content to make embedded ML accessible. Comfortable across low-level build systems (CMake/flatbuffers) and TFLite/TFJS export pipelines, he brings a maker’s background from Shenzhen’s hardware ecosystem and a formal BE in Electronics from Beihang University.
8 years of coding experience
5 years of employment as a software developer
Trinity college kandyBachelor of Engineering - BE, Electronics and Information Engineering, Bachelor of Engineering - BE, Electronics and Information Engineering at Beihang Universityhigh school, mathematics,physics,chemistry, high school, mathematics,physics,chemistry at Royal college colomboPost-Baccalaureate Diploma, Computer and Information Systems, Post-Baccalaureate Diploma, Computer and Information Systems at Douglas College
Sinhalese, English, Chinese
Stackoverflow
21reputation
350reached
0answers
1question
Github Skills (19)
pytorch10
python10
tensorflowjs10
machine-learning10
onnx10
deeplearning-ai10
deep-learning10
computer-vision10
tflite10
nvidia-jetson9
arm9
openvino8
nodejs6
javascript6
electron6
Programming languages (13)
Github contributions (5)
ultralytics/ultralytics
Mar 2023 - Apr 2025
Ultralytics YOLO11 🚀
Role in this project:
ML Engineer
ML EngineerContributions:34 reviews, 106 PRs, 423 pushes in 2 years
Contributions summary:Lakshantha primarily contributed to the export and optimization of the Ultralytics YOLO model across various platforms and hardware. Their work includes fixing issues related to flatbuffer and cmake requirements for specific architectures like ARM64 and NVIDIA Jetson. They also optimized exports for TensorFlow.js and TFLite, and addressed compatibility issues. Moreover, the user improved the `is_jetson` function and implemented features such as including `data` argument for the Sony IMX500 export, as well as adding a `device` argument to all the exports, which is crucial for hardware acceleration.
pytorchdeep-learningyolov8object-detectiononnx
lakshanthad/hello-world
Jan 2020 - Dec 2024
This is a test repo
Contributions:1 release, 2 PRs, 4 pushes in 4 years 11 months
testingtest-repo
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.