Thijs Gelton
Deep Learning Engineer at Oddity.ai- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
-
Nederlands Native or bilingual proficiency
-
Engels Professional working proficiency
-
Duits Limited working proficiency
Topline Score
Bio
Credentials
-
MongoDB for Developers
MongoDBMar, 2017- Oct, 2024
Experience
-
Oddity.ai
-
Netherlands
-
Software Development
-
1 - 100 Employee
-
Deep Learning Engineer
-
Jan 2023 - Present
Responsible for the development pipeline of the violence detection algorithm, bridging the realms of industry and academia. Using Python for data collection, cleaning and labelling of videos and both PyTorch and TensorFlow for model development using the latest computer vision techniques. Responsible for the development pipeline of the violence detection algorithm, bridging the realms of industry and academia. Using Python for data collection, cleaning and labelling of videos and both PyTorch and TensorFlow for model development using the latest computer vision techniques.
-
-
-
Radboudumc
-
Netherlands
-
Hospitals and Health Care
-
700 & Above Employee
-
Research Intern RadboudUMC
-
Feb 2022 - Dec 2022
Researched the possibilities of AI-assisted Detection of Endometrium Carcinoma in Pipelle Biopsies. This project resulted in a grade of 9.5 and graduating Cum Laude. • Developed a weakly supervised multiple instance learning algorithm that could highlight areas of importance using self-attention (project: https: bit.ly/3Vm8pOS, demo: https://bit.ly/3VCsAYC) and set up a challenge on data of our reader study to stimulate open science (challenge: https://bit.ly/3UlMeHh). This is being adopted by the National Institute of Health (NCI) on their online platform. • Developed a multi-branch multi-resolution architecture as an extension to the nnU-Net framework to inspect the possibility of semantic segmentation on Pipelle sampled endometrial tissue (GitHub: https://bit.ly/3uvM9GB). Show less
-
-
-
4net Interactive (part of 4NG)
-
Netherlands
-
Software Development
-
1 - 100 Employee
-
Junior Dotnet Developer
-
Oct 2020 - Dec 2022
Solving defects in business critical systems of our clients. This requires full stack development (ASP.NET MVC, React, Vue.js, etc.) and DevOps knowledge. Solving defects in business critical systems of our clients. This requires full stack development (ASP.NET MVC, React, Vue.js, etc.) and DevOps knowledge.
-
-
-
4NG
-
IT Services and IT Consulting
-
Junior Dotnet Developer
-
Oct 2020 - Dec 2022
-
-
-
PossibilIT BV
-
Netherlands
-
Software Development
-
Dotnet Developer
-
Feb 2019 - Oct 2020
Working on project basis to create middleware applications between SalesForce and Dynamics CRM and subscription software like Zuora and solving a plethora of data engineering tasks. This required .NET Core, Azure Cloud Computing, Docker and Kubernetes knowledge. Working on project basis to create middleware applications between SalesForce and Dynamics CRM and subscription software like Zuora and solving a plethora of data engineering tasks. This required .NET Core, Azure Cloud Computing, Docker and Kubernetes knowledge.
-
-
-
Betabit
-
Netherlands
-
IT Services and IT Consulting
-
100 - 200 Employee
-
Graduate Internship
-
Aug 2018 - Jan 2019
Researched the possibilities within the Azure domain to create and deploy a Computer Vision object detection model (ResNet, Azure Cognitive Services, SVM w/ HOG descriptor). The best fit model was deployed in a Angular web application with a .NET Core backend. Its purpose was to detect and index LEGO and show a list of possible sets to build. Researched the possibilities within the Azure domain to create and deploy a Computer Vision object detection model (ResNet, Azure Cognitive Services, SVM w/ HOG descriptor). The best fit model was deployed in a Angular web application with a .NET Core backend. Its purpose was to detect and index LEGO and show a list of possible sets to build.
-
-
-
Shopping Minds
-
Netherlands
-
Technology, Information and Internet
-
1 - 100 Employee
-
(Full time) Junior Fullstack Developer
-
Jan 2017 - Jun 2018
Worked as a full stack developer on customer data collection on client websites. This data would be used for customized campaigns and promotions. I worked on automating the process of creating campaigns based on this data and later co-developed the Data Management Platform using Python (Flask/Web2py), MongoDB (M101P certified), JavaScript/jQuery). Worked as a full stack developer on customer data collection on client websites. This data would be used for customized campaigns and promotions. I worked on automating the process of creating campaigns based on this data and later co-developed the Data Management Platform using Python (Flask/Web2py), MongoDB (M101P certified), JavaScript/jQuery).
-
-
Education
-
Radboud University
Master of Science - MS, Artificial intelligence -
Hogeschool Utrecht
Bachelor's degree, Informatica -
R.S.G. Pantarijn
V.W.O.