Welcome To My Portfolio
Pomodoro
A productivity timer
Pomodoro is a productivity timer built with React and TypeScript. The project is open source and can be found on GitHub. In this project it was a goal to experiment with colors and understand how colors interact with eachother to better understand how to design in the future. I decided to experiment with a mono-chromatic color scheme.
React
TypeScript
Pomodoro
Timer
Productivity
Next.js
Tailwind
Full Stack
Server-side rendering
React
TypeScript
Pomodoro
Timer
Productivity
Next.js
Tailwind
Full Stack
Server-side rendering
University projects
Bachelor's project
Digitization of whiteboards and papers in real time using a mobile device
In this paper we combine several techniques within the computer vision area in order to develop a complete system for capturing a whiteboard or paper and make it presentable for an audience in real time. Our system is implemented on a regular smartphone hence enabling users to do remote presentations without requiring any extra tools. We divide our method into five major steps that take care of a specific part of the problem. We end up with a system capable of capturing drawings on both paper and whiteboards in different conditions with the possibility for further improvements.
Whiteboard Digitalization
AI
Research
Deep Learning
Corner Detection
Canny Edge Detection Algorithm
Image Processing
Image transformation
Semantic Segmentation
DeepLabV3
Atrous Spatial Pyramid Pooling (ASPP)
Binarization
Change Detection
Java
Python
OpenCV
Tensorflow
Whiteboard Digitalization
AI
Research
Deep Learning
Corner Detection
Canny Edge Detection Algorithm
Image Processing
Image transformation
Semantic Segmentation
DeepLabV3
Atrous Spatial Pyramid Pooling (ASPP)
Binarization
Change Detection
Java
Python
OpenCV
Tensorflow
Automated playtest with AI
Calculation of optimal play strategies
The purpose of this paper is to create an AI based program that can partially automate certain parts of the playtesting phase in the game development industry. The program is to reduce the time and cost of the playtesting phase. The solution program developed is chosen to be based on a neural network and a genetic algorithm. Furthermore a snake game is developed in order to test the programs functionality. Two methods of playtesting are successfully automated. However it is concluded that the program can only be considered a proof of concept. This is due to the programs lack of ability to handle complexity as well as the amount of implementation required should the program be effectively used on another game. Furthermore if the program was ever to be used on more complex games then the neural network and the genetic algorithm would need to be significantly expanded and large amount of computing power would be necessary.
Game Development
Test games
AI
Playtest with AI
Snake
MoSCoW Analysis
Neural Network
Genetic Algorithm
Game Development
Test games
AI
Playtest with AI
Snake
MoSCoW Analysis
Neural Network
Genetic Algorithm
Map Matching and Weight Completion
Complex back-end development
The following report is centered around the design and implementation of a Map Matching and Weight Completion service for the aSTEP platform. The goal of the Map Matching service was to provide the necessary data for the Weight Completion service to be functional. Both services were successfully implemented and integrated on the aSTEP platform. During the semester collaboration was a major focus with the Map Matching service being developed in collaboration with Group SW505. Furthermore all aSTEP groups of the semester collaboratively helped improve and maintain the aSTEP platform by updating its documentation as well as optimizing the server architecture. Finally the agile method Scrum was used for the development of the project which is also documented in this report.
Map Matching
Weight Completion
AI
Kubernetes
Docker
Java
Python
JavaScript
Scrum
Collaboration
Mega project
Unit test
UI
Website
Graph Convolutional Neural Network
aSTEP
CI/CD
Map Matching
Weight Completion
AI
Kubernetes
Docker
Java
Python
JavaScript
Scrum
Collaboration
Mega project
Unit test
UI
Website
Graph Convolutional Neural Network
aSTEP
CI/CD
