Within the context of the embedded AI lab, we are working on several projects. If you are a bachelor or master student and interested in participating in such a project, please refer to the Available Assignments page.
Running projects
- [Internship] Research and Implementation of Embedded Facial Recognition Systems
This is a confidential project in collaboration with 20Face.
- [Internship] Anomaly Detection based on Hydraulics Pressure Data of an Autonomous Excavator
This is a confidential project in collaboration with Fraunhofer IOSB.
- [MSc Thesis] Exploring Hardware-Aware Neural Architecture Search for Cracked Egg Detection
- [MSc Thesis] Exploring Application-Specific Instruction Set Processors (ASIP) for YOLO Algorithms
- [MSc Thesis] Image Analysis for Quality Assurance in Seeding Patterns at GROWY
Project Context: At Growy, we are dedicated to optimizing agricultural practices through advanced technology. This project is a key part of our strategy to incorporate state-of-the-art image analysis into our seeding process.
Project Overview: The goal of this master project is to develop a quality assurance station that employs cameras to inspect the seeding patterns created by our new seeding machines. This system will verify if these patterns align with Growy's quality standards and will decide if the sown gutter is allowed to enter our farm
Finished Projects
- [BSc Thesis] Exploring Layer-Specific Quantization in CNN-Based Selective Sweep Detection
This bachelor's assignment explored the effectiveness of layer-specific quantization within the field of population genetics. This study showed that reducing the layers of our existing convolutional neural network to a 2-bit configuration had minimal effect on the accuracy of the network. The study also showed some layers could even be reduced to 1-bit in some circumstances. These results stimulate more research on the topic and the development of specialised hardware that allows for the 2-bit neural network.
The paper can be read here