This page lists previous theses written in the AIML group, including links to the published thesis documents (as far as currently available).
Bachelor's Theses
To be published:
2024
Dennys Huber: OpenMax with Clustering for Open-Set Classification
Pablo Tanner: Automatic Facial Emotion Recognition through Information Fusion
2023
Lars Bösch: Suppressing the Extraction of Features from Unknown Samples
Master's Theses
To be published:
2024:
Christian Berger: Spatio-Temporal Time Series Forecasting with Long Short-Term Memories and Graph Neural Networks
Juan Diego Bermeo Ortiz: Test Time Adaptation with Denoising Diffusion Probabilistic Models for Medical Image Segmentation
Melih Catal: Does Adversarial Training Improve Adversarial Stability?
Sarah Feuz: Effective Negative Samples for Open-Set Classification
Michèle Fundneider: Demographic Bias in Face Recognition: Evaluating WERM Fairness Metric and Balancing Strategies
Simon Giesch: Improvement of Open-Set Classification through Adequate Loss Weighting of Negative Samples
Dean Heizmann: Generating Negative Samples for Open-Set Classification
Silvan Kübler: Open-Set Classification with Ensembles of Binary Classifiers
Dongyi Lang: Combined Out-of-Distribution Detection and Open-Set Classification
Nan Li: Interpretable Machine Learning for Drunk Driving Detection
He Liu: Open-Set Classification for Human Activity Recognition Using Wrist-Worn Accelerometers
Agnar Pétursson: Visualization of Image Similarities in Face Recognition Systems
Melanie Salzer: Adversarial Stability of Open-Set Deep Networks
Le Hoang Minh Trinh: Mapping of Socio-economic Development using Satellite Imagery; Addressing Data Imbalance in Regression Models And Its Effects on Visualizations
Jun Tu: End-to-end Open-set Recognition using Probability of Inclusion
Joël Watter: Face Recognition Based on Facial Attributes in Degraded Images
2023
Yuning Yu: Learning Semantics of Classes in Image Classification: Modifying the loss function
