Ajay Balasubramaniam
Deep Learning Engineer at NavInfo Europe B.V.- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
-
Tamil Native or bilingual proficiency
-
English Full professional proficiency
-
Japanese Elementary proficiency
-
Dutch Elementary proficiency
Topline Score
Bio
Credentials
-
Machine Learning
CourseraJun, 2016- Nov, 2024 -
DAT208x: Introduction to Python for Data Science
edXFeb, 2016- Nov, 2024 -
Diploma in Embedded System Design (DESD)
TCILJan, 2012- Nov, 2024
Experience
-
NavInfo Europe B.V.
-
Netherlands
-
IT Services and IT Consulting
-
1 - 100 Employee
-
Deep Learning Engineer
-
Jun 2020 - Present
Real-time deployment of State-of-the-art deep learning solutions Worked on several challenging computer vision topics from understanding new research till deployment and maintenance. * Classification * Object detection * Semantic segmentation * Multi-task models * Online continual learning * Multi-modal LLMs Deployment on: + NVIDIA desktops, servers + AWS EC2 + NVIDIA edge devices + Android mobile phones Real-time deployment of State-of-the-art deep learning solutions Worked on several challenging computer vision topics from understanding new research till deployment and maintenance. * Classification * Object detection * Semantic segmentation * Multi-task models * Online continual learning * Multi-modal LLMs Deployment on: + NVIDIA desktops, servers + AWS EC2 + NVIDIA edge devices + Android mobile phones
-
-
-
Thermo Fisher Scientific
-
United States
-
Biotechnology Research
-
700 & Above Employee
-
Deep Learning Researcher
-
Feb 2019 - Feb 2020
Under 60 seconds semantic segmentation training!! Smart-microscopes: Deep learning based semantic segmentation on life-science tomograms. Ultra fast training. Human-in-the-loop active learning paradigm to label whole 3D tomograms in minutes which otherwise take hours to days to complete. Under 60 seconds semantic segmentation training!! Smart-microscopes: Deep learning based semantic segmentation on life-science tomograms. Ultra fast training. Human-in-the-loop active learning paradigm to label whole 3D tomograms in minutes which otherwise take hours to days to complete.
-
-
Education
-
Eindhoven University of Technology
Master's degree, Embedded Systems -
SRM University
Bachelor of Technology - BTech, Computer Science