Review of the Project Day of the Faculty of Computer Science at Augsburg University of Applied Sciences in the summer semester 2024
On June 19th, 2024, the largest Project Day of the Faculty of Computer Science at Augsburg University of Applied Sciences took place. The event provided an impressive platform for students to present their innovative projects and promote the exchange between science and practice. Enthusiasm for algorithms, data and digital technologies was at the forefront and enabled participants to implement creative ideas, develop innovative applications and optimize digital processes. Whether it was the development of software and apps, the analysis of data, research into artificial intelligence or the functioning and control of drones - the project day offered a wide range of insights and information on the most diverse areas of computer science.
Students from the Bachelor's and Master's degree courses presented the results of their semester projects live in a total of 49 teams. Thanks to cooperation with partners from business and society, the projects were able to establish a direct practical relevance. We also took part with a project in the field of additive manufacturing, which was successfully completed by a team of three students.
CADdent project: Effective recognition of 3D-printed objects
As part of our project, we worked on an automated assignment of individualized dental technology components. This is because once production is complete, the loose components, which have been grouped together in a building cicle, have to be manually assigned to their orders. When automating this process, the uniqueness and intricate size of the components poses a major challenge for conventional object recognition methods.
The project team tested various techniques ranging from classic image processing to 3D-based approaches. Image-based deep few-shot learning proved to be the most promising method. Specifically, a small amount of synthetic image data is generated for each manufactured job, whose abstract features are extracted by CLIP and used to train a simple feed-forward classifier. By prototyping across multiple camera perspectives and using transductive inference, we can achieve a high degree of accuracy when assigning webcam images of the components to their customer orders.
Winning team of the project day
The participating teams and visitors voted for the best projects in a poll. First place went to the project "EcoHeroes V2.0 - VR for sustainable textile recycling". This VR game is designed to teach children about textile recycling in a fun way
As some of our employees studied at Augsburg University of Applied Sciences themselves, we were particularly pleased to be able to provide and present the main prize to the first-place winners. The winning duo Lukas Diemer and Domenic Schäfer can look forward to a voucher worth €1,000 for rapid prototyping products. They will also receive a company tour at our main facilities in Augsburg.
The winners Lukas Diemer and Domenic Schäfer, together with the Dean Prof. Dr. Alexander von Bodisco and our software developer Simon Kunzmann from CADdent.
Have we aroused your curiosity?
CADdent GmbH offers students the opportunity to complete final theses, internships and practical semesters at our premises. If you are interested in working with us, please get in touch! We look forward to working with you on exciting projects and shaping the future of digital technologies.