Open Student Projects, Bachelor and Master Theses at the Institute of Theoretical Computer Science

We are looking for highly motivated students who want to participate, during their Bachelor/Master Project and/or Thesis, in research on areas of machine learning, deep learning, and brain-inspired paradigms for computing and learning. Concete topics for Bachelor theses (B), seminar projects (P), Master theses (M) are listed below. Topics can be specialized further in discussion with the student, in order to accomodate prior knowledge and interests.

Project Topic Type Contact
Backpropagation with low-precision gradient accumulation B/P/M Robert Legenstein
Exploring gradient-based optimization techniques for deep learning B/P Robert Legenstein
Deep learning for classification out-of-context: How much do deep networks rely on backgrounds? B/P/M Ozan Özdenizci
Robust generative modeling with variational autoencoders B/P/M Ozan Özdenizci
Exposing adversarial vulnerabilities of DNNs in practice B/P Ozan Özdenizci
Robust pruning of DNNs by harnessing adversarial decisions B/P Ozan Özdenizci
Unraveling the DNN black-box: explaining decisions of multi-task deep learning architectures B/P Ozan Özdenizci

Institute of Theoretical Computer Science
Inffeldgasse 16b/I
8010 Graz

Phone: +43 (0) 316 / 873 - 5811
Fax:       +43 (0) 316 / 873 - 105811

Univ.-Prof. Dipl.-Ing. Dr. techn. Robert Legenstein