Bachelor, Master and Doctoral School Projects

Counterfactual explanations  (CEs) aim to alter output labels with minimal disruption. While adversarial attacks do it regardless of the nature of these changes, CEs restrain its attack to remain in the data manifold. To this end, current methods utilize diffusion processes ([1],[2]) for direct optimization, which, although effective, require significant memory and computational resources for (…)

Explainable AI (XAI) seeks to tackle the opacity of deep neural network decisions. Moving beyond the conventional focus on 2D imagery ([1],[2]), our research provides the first method to provide Counterfactual Explanations (CEs) for 3D point cloud classifiers. Specifically, we propose for the first time a method for creating counterfactuals with explicit semantic guidance.

Recent 3D reconstruction methods learn a 3D representation of the scene from a set of 2D images. While these representations are able to model the scene with a very high visual accuracy, they make edition cumbersome and they are unable to infer parts of the objects which are unseen in the initial images. In particular, (…)

Demonstrate spacecraft pose estimation algorithms running on small edge devices.

Modern 3D reconstruction techniques rely on pixel consistency across images to recover 3D geometry from 2D images. These methods include NeRFs and Gaussian Splatting, which are the most accurate techniques in current state-of-the-art. However, they struggle on inputs with little contrast, objects which are transparent or reflective, and views which are too few or with (…)

This project aims to enhance object detection and tracking algorithms by integrating scene geometry into the detection process. Utilizing mesh data of the scene and video footage from one or multiple cameras, the project seeks to develop a detection system that can leverage this additional information to improve accuracy and efficiency.

We live in a world full of manufactured objects of ever-increasing complexity that require clever engineering to be functional. Today, individual parts of these objects are often optimized separately, control is rarely accounted for at design time, and much manual tinkering is required. These shortcomings can largely be ascribed to the composite nature of industrially-designed (…)

In particular we are targeting the task of 3D tracking and long-term tracking of people. The first step toward that goal will be to build a multi-camera dataset suitable for those tasks. Initially the student will design a multi-view annotation tools that will leverage camera calibration to minimize annotation cost.

Develop a comprehensive framework for generating near-photorealistic images that accurately represent deformable objects and changing materials.

Computer Aided Engineering (CAE) is at the core of modern industrial engineering and manufacturing. However, the current CAE applications suffer from significant time and human resource expenses. Our goal is to leverage deep learning techniques to automate the CAE process and reduce the R&D costs for the industry.

The project presented here aims to address these limitations by fine-tuning a Large Language Model (LLM) to the specific cultural and linguistic characteristics of the Canton de Fribourg.

The goal of this project is to develop a tool that allows the insertion of realistic synthetic observations of space objects into astronomical images.