SCS - Studienarbeiten

Der Golfschwung gilt als einer der komplexesten Bewegungsabläufe aller Sportarten. Für den Amateurgolfer besteht die Schwierigkeit darin, bei einem Fehlschlag den Fehler in seiner Bewegung zu erkennen und entsprechend zu korrigieren.

Ziel 1: Entwicklung einer Physik-Engine mit geeigneten Parametern für die Berechnung der Flugbahn des Golfballs.

Ziel 2: Automatisierte Analyse des Bewegungsablaufs des Golfspielers und gezieltes Erkennen fehlerhafter Muster. Dazu soll ein geeignetes Verfahren ausgewählt und umgesetzt werden, wie z.B. Vergleichen mit Bewegungsabläufen von Profis.

Der Simulator soll mit Low-Cost COTS Komponenten aufgebaut werden.

Continue reading “Students Internships AI-based Golf-Coach with Automated Swing Analysis”

SCS - Studienarbeiten

Stereo Vision allows accurately measuring scenes in 3D – provided that the cameras are well calibrated (camera pose, intrinsic parameters). Especially in harsh industrial or automotive contexts, the mechanical vibrations, temperature variations, and material fatigue affect opto-mechanical properties of the stereo-rig and its measurement accuracy.A jointly developed intelligent door opener (embedded stereo-system: can detect, discriminate (humans, vehicles), and track 3D objects in real-time. This enables adequate door openings, e.g. suppressing cross-traffic and accounting for the object’s height to minimize energy loss in cooling storage applications. Because calibration drifts can cause e.g. inappropriate openings/closing, periodic or continuous recalibration of the stereo-rig is advantageous.The goal of this thesis is to explore, test and analyze different methods and algorithms for continuous stereo-vision calibration. The solution has to determine the extrinsic and intrinsic camera parameters without any markers or reference objects. Its performance will be measured mainly in overall accuracy, but also calculation time and memory consumption.

Continue reading “Self-Calibration for Embedded Stereo Vision System”

SCS - Studienarbeiten

Today’s Machine Learning (ML) success is often limited by a lack of labelled ground truth (GT) data to train the models. This is especially true for applications in medical imaging.Transfer learning (TL) is a state-of-art ML-technique that can be useful to overcome this problem for similar yet distinct tasks. TL tries to apply knowledge or models gained from a task to a second one. In the field of image recognition, TL-methods for deep neural nets re-use parameters trained on source domain data except for the output layer (see image on the right).In medical optical coherence tomography (OCT) of the eye only a limited set of GT-labeled images exists. In addition, there are several device manufacturers and imaging approaches. Thus, transfer learning could 1) enable an easier addition of new device types or imaging techniques, 2) improve the quality of available segmentations even with limited data.

Continue reading “Transfer Learning for Image Segmentation using Convolutional Neural Networks”

SCS - Studienarbeiten

Training effective Artificial Intelligence (AI) algorithms today often requires large amounts of ground truth data. Typically, this is a laborious, costly and time-consuming process often requiring manual adjustment. These problems can be overcome by combining AI with Virtual Reality (VR): an emerging technology with various applications in medicine, training, business and simulation. VR creates artificial environments that often resemble our real world.This project aims at exploring the potential of VR as 3D ground truth generator for state-of-the-art Machine Learning (ML) and Deep Learning (DL) Algorithms.

Continue reading “Virtual Reality as 3D Ground Truth Generator for AI, Machine Learning and Deep Learning”

The World Health Organization (WHO) assumes that over 25% of the medicine sold in developing countries and about 50% of medicine ordered online are fake. According to WHO’s estimations, the death toll in Africa alone amounts to 100’000 per year due to fake medicine. As a reaction to these figures, numerous governments worldwide are releasing more restricting regulations for pharmaceutical products (packaging, seals, verification etc.) and are obliging the industry to ensure the traceability of corresponding products.

The packaging plant manufacturer Gerhard Schubert GmbH considers it normal that his facilities ensure the total traceability of each product in every packaged unity and every packaging level (i.e. clusters, boxes and pallets) and thus meet the regulatory requirements of customers, as well as national and international requirements.

It is crucial that the track and trace functionality, as well as the implementation of ERP functions does not negatively impact the output and performance of the production facility. Moreover, it must feature the flexibility to be adapted by the customer when requirements change and must show a complete audit-trail.

Thanks to the modular architecture implemented by SCS, the various regulatory requirements concerning labelling and verification can be configured and expanded by the customer independently. Also, ERP systems of various manufacturers can easily be implemented. The integrated audit-trail ensures the required tracking of all relevant changes made to the system.

Applied Technologies: Microsoft .NET/C#, SQL-Server, WCF