We constantly offer student projects within the following fields of computer vision, pattern recognition and machine learning. The only qualifications you have to bring along are

• interest in Computer Vision and
• basic programming skills.

We usually develop using either Python or C++ - so, basic knowledge in either of those would be beneficial.

Typically, each project can be scaled according to your requirements. So if you're interested in a MSc Thesis, Seminar/Master project, or a BSc Thesis, just contact us.

• How can we ensure to detect all objects to prevent accidents?
• How can we verify an algorithm's correctness to allow self-driving cars on Austrian roads?
• Which sensors can be fused to enable robust visual perception in real-time?

If you're interested in experimenting with state-of-the-art approaches to improve visual perception of autonomous vehicles, contact us!

• How can we teach computers to detect objects in images?
• How does a computer tell a cat from a human?
• Does 3D vision (e.g. LIDAR, Stereo, ToF) simplify object detection?

 If you're interested in experimenting with state-of-the-art object detectors or getting first-hand experience with machine learning approaches, contact us!

• How can we recognize and interpret hand gestures or body poses?
• How can we leverage gesture recognition for HCI applications, such as gaming, action analysis or sign language recognition?
• How can we learn a robust model from few training samples?

If you're interested in experimenting with recent machine learning approaches to improve hand and/or body pose estimation, gesture recognition or automatically generating realistic training data, contact us!

• Have you ever wondered which algorithms enable computers to see?
• How can we improve artificial intelligence?
• What are open problems for cognitive computing systems?

If you're interested in experimenting with state-of-the-art machine learning approaches to solve computer vision problems, tackle pattern recognition tasks or build applications that see, contact us!

• How can we localize objects throughout a video sequence?
• What is the purpose of motion analysis and object trajectories?
• How can we predict the intent of a human by observing gait?

If you're interested in automated video analysis, understanding what's going on in a scene, or just want to get first-hand experience with tracking algorithms, contact us!

• How can we tell one person/bird/car from the other?
• How can we recognize and identify previously seen individuals?
• What are the caveats for image retrieval?

If you're interested in person re-identification, want to know how face unlock works on your smartphone, or are curious about fine-grained classification, contact us!

• How can we extract semantically meaningful information from images?
• How can we distinguish overlapping instances of the same semantic category?
• How can we interpret the semantic data to understand what is going on?

If you're interested in improving scene understanding algorithms, experiment with activity understanding approaches or are keen to find out, where semantic information is required in real-world applications, contact us!