Home | Friso Verweij

Who am I?

Hello! I'm Friso, Master of Science in artificial intelligence

In 2022 I finished the master's program Artificial Intelligence at the University of Amsterdam, which is, according to program director at least, the most technical AI study in Europe. My focus within the program was machine learning, deep learning and computer vision. I have also followed courses about natural language processing, reinforcement learning and information retrieval. During my thesis I specialised in computer vision, creating a complete pipeline that estimates the position and intensity of light sources in images that are used to insert and render virtual 3D objects.

Before studying at the University of Amsterdam, I followed the bachelor's program Kunstmatige Intelligentie (Artificial Intelligence) at Utrecht University. In contrast to the technical nature of the master's program, Utrecht offered a much more interdiciplinary view on artificial intelligence. Besides learning datastructures and algorithms, I also learned about the psychological and philosophical nature of AI. During my time at Utrecht University I had the wonderful opportunity to be given the role of Teaching Assistent for a course related to multi-agent systems, game theory and mathematics.

I am a very curious person. Learning, growing and progressing is something I very much enjoy. I love to code and to learn more about software and it's creation. I have a big passion for games and the creation of games. Anything that is creative and is related to computers I eventually lay my hands on, from video editing, creating games, and webdesign to photoshop.

Interests

Computer Science

Formula 1

Game Development

Web Development

Computer Graphics

Videogames

Boardgames

Virtual Reality

Playing instruments

Education

Master of Science Artificial Intelligence
University of Amsterdam
2020 - 2022

After finishing my bachelor's, I wanted to get a better understading of the technical side of AI, focusing more on machine learning and it's derivatives. The following two years I studied Artificial Intelligence at the University of Amsterdam. Some of the courses I have followed:

Machine Learning 1
Machine Learning 2
Computer Vision 1
Computer Vision 2
Deep Learning
Natural Language Processing
Information Retrieval
Reinforcement Learning
Data-Driven Business Innovation and Entrepeneurship

A large portion of the master's program is the master's thesis. In total the thesis will take at least 8 months. For my thesis, I created IllumiNet and IllumiSet. IllumiNet is a neural network that uses RGB-D images to learn where light sources are positioned in an image and what their intensity is. These retrieved light sources can then be used to insert virtual object into the image. To train the network, no existing dataset met my needs. As a result, I created IllumiSet, a large dataset of synthetic RGB-D images, along with intrisic images, of indoor scenes created using Blender. I was supervised by the company 3DUniversum, a spin-off company from the University of Amsterdam.

Bachelor of Science Kunstmatige Intelligentie
Utrecht University
2017 - 2020

Before my master's programme, I followed the bachelor's programme Kunstmatige Intelligentie (Artificial Intelligence) at Utrecht University. In contrast to the master's programme, the bachelor had an interdiciplinary focus. Besides learning to create smart algorithms, there was an emphasis on psychology, ethics and philosophy. This provided me with a strong foundation to continue my studies at the University of Amsterdam. In the final year of the bachelor, I followed the minor in Computer Science. The following are some of the courses I have followed (translated into english when applicable):

Introduction to Cognitive Science
Introduction to Logic
Logic for AI
Mathematics for AI
Introduction to Adaptive Systems
Datastructures and Algorithms for AI
Computational Linguistics
Experimental Methods and Statistics
Machine Learning
Intelligent Systems
Computational Intelligence
Sensation and Perception
Computer Architecures and Networks
Concurrency
Modelling and System Design
Graphics

I wrote my thesis about neuroevolution, where I created and examined an evolutionary algorithm that was able to create fully trained neural network architectures for multi-class classification tasks, without modeling to a specific dataset. This algorithm could therefore be used to create a neural network by itself, regardless of dataset.

Working experience

Computer Vision Research Intern
3DUniversum
2021 - 2022

For my master's thesis, I did a research project at 3DUniversum, a spin-off company from the University of Amsterdam that provides AI services. I created IllumiNet a neural network that learns to recover the position and intensity of light sources in RGB-D images for object insertion. While the company provided supervision, the project was primarily hands-off on their part. This gave me a lot of freedom to design and implement my ideas, resulting in a thesis I am very proud of.

Teaching Assistent:
Inleiding Adaptieve Systemen
Utrecht University
2022

During my bachelor, I had the wonderful opportunity to be a asked to become a teaching assistent (TA) for the course Introduction to Adaptive Systems. This course was an introductory course for first-year students covering math, evolutionary algorithms, multi-agent systems, game theory and of course a lot of programming. Most of the contents of the course are algorithms and principles inspired by nature. As these processes have had thousends of years to evolve and perfect itself, a lot of them seem to act intelligently. Think for example about a colony of termites that use pheromones to communicate the location of food, creating a very efficient traveling network. By mimicing their behaviour, we create a very efficient network between, for example, train stations.

While I was the TA, I had the responsebility over about 30 students. This involved explaining the concepts, answering questions, helping with programming and grading their work. Even though the course only lasted 2.5 months, it was an incredible experience. I thoroughly enjoyed teaching students, working for a good cause and being a leader figure.

Runner
De Beren Restaurants
2019 - 2020

As life was getting expensive, I started working as a runner at "De Beren", a restaurant chain. I was tasked with delivering as many drinks as fast as possible with the largest possible smile. While it was nice to be talk and serve so many people, the restaurant was overstaffed, resulting in few hours. Eventually I left this job as I started working as teaching assistent for Utrecht University.

Sales Hifi and TV
MediaMarkt
2017 - 2018

For over a year I worked as HiFi advisor to customers at MediaMarkt. Because of my interest in electronics, I enjoyed telling other people about them and inspire them. For me, it was not a matter of making the most sales, but to make the most customers happy. Because a stereo installation might not be interesting to everyone, I saw it as a challenge to excite everyone about it. Seeing grumpy people change their frown to a smile and people remembering my name was the best part of the job. Later I also started selling TV's, which is also easier to excite people about. In the end I left the company because I did not like their targets. Instead of focusing on selling the products and making customers happy, it was more important to lure as many people to Ziggo and energy provider stands. I thoroughly disliked is, as the smile I just brought on the customers face vanishes immediately.