- Linz Osterreich
- [email protected]
The primary research goal of my master’s thesis project was to design a theoretical computer vision system capable of detecting hand gestures in real time using a regular smartphone camera. Practically, I had to built a prototype that, using a static camera, was able to detect hand gestures in real time. All the practical machine learning parts (SVMs, kNN) of the project were based on the OpenCV (for Windows) library. In the end, the resulting system was able to detect various types of hand gestures and displayed commercially required speed to accuracy ratios.
From a software development point of view, in my three years stay at Augmensys, I mainly worked as an Android developer on the client side of the company’s core project. As such, I was involved in all main development areas: (1) adding new features, (2) improving the quality/functionality of the existing features, (3) quality assurance (i.e., unit testing, system testing and continuous integration). With regard to precise activities, I was involved in all the areas of the application kernel: user interface, augmented reality, Google Maps R , REST protocol, certificates, databases, etc.
Starting with May 2014, my day-to-day activities switched to Windows 8 development and I began implementing new features for the win8.1 version of our application.
My main task involved formalization of client specifications and the rapid prototyping of a fairly complex customer management application based on these specifications.
Teaching Assistant for one semester, I was responsible with the 101 Mathematical Analysis seminars at the University of Agronomic Science and Veterinary Medicine Bucharest. The covered exercises included basic topics like sequences, series, continuity, and partial derivatives etc.
This course is about learning algorithms and data structures, focused on performance of applications. The language used in this course was java. The topics included are, union-find algorithms, sorting algorithms quicksort, mergesort, heapsort, basic data types stacks, queues and bags, binary search trees, red-black trees, hashtables and symbol table algorithms.
The general field of my Master’s Thesis is computer vision with the construction of a gesture recogni- tion system as the main goal. A systematic trial and error approach finally resulted in two possible designs: (1) a model based on a hybrid SVM / kNN classifier and (2) a model that used a background subtraction algorithm. A comparative analysis yielded that the former can deal with broader classifi- cation problems, while the second approach was able to accurately discriminate between background and foreground (i.e., a moving hand) much faster in real time when the camera was static
I mainly studied and conducted research in algebraic fields with a focus on number theory. The thesis is titled “Generalizations of Pell Equations”.
From 2005 to 2008 I studied for three years according to a combined mathematics and computer science curricula. The primary field was mathematics and accounted for ≈ 60% of the course load. The title of my Bachelor’s degree in Mathematics is “Ring Extensions”.
In 2010 I also received my BSc in Computer Science. The thesis is related to increasing security in asp.net websites. The concrete examples I worked with were related to a web store specialized in visualizing music online and an online store for goods. In both cases, customers were able to edit, delete and add data to their personal virtual shopping carts.
Prof. Dr. Bruno Buchberger - Tel. +43 (0)664 421 1646
Prof. Dr. Josef Küng - Tel. +43 (0)664 6024 68890