Teaching & theses

In the spring term 2026, we are teaching:

• Mobile Health and Activity Monitoring 2026 master course
• Adaptive User Interfaces through Machine Learning 2026 bachelor seminar
• Egocentric Computer Vision: Perception and Understanding from the First-Person View 2026 master seminar
• Reinforcement Learning for Understanding and Modeling Human Behavior 2026 master seminar

Available student thesis projects

6 projects are available (via 6 master’s theses and 5 semester theses). Currently, all bachelor’s theses have been assigned. The list below is up-to-date.

Please submit your CV and transcripts through SiROP. In case of questions, please contact us via SiROP.

• Making AI agents robust for using imperfect tools

  Tool-augmented LLM agents often rely on tools that are usually correct but can occasionally return plausible, incorrect outputs. In this project, we will mitigate these silent errors through a cost-aware strategy that combines cross-tool consistency checks with targeted verification at high-risk steps.

  more...
  
  • contact: Tommaso Bendinelli
  • semester thesis
  • master's thesis

  • Agent reliability
  • Tool-augmented LLM Agents
  • Robust reasoning

• Benchmarking agentic reasoning robustness on time series

  We will create a research benchmark for evaluating AI agents on decision making from time series. Our benchmark will integrate simulations that expose realistic reasoning failures and edge cases. We will evaluate open-source agentic systems using this benchmark on robustness, reliability, and cost-aware performance.
  The project combines benchmark engineering, controlled experimentation, and empirical analysis to identify which agentic strategies are reliable in realistic time-series decision settings.

  more...
  
  • contact: Tommaso Bendinelli
  • semester thesis
  • master's thesis

  • ⁠Time-series reasoning
  • Agent reliability
  • Benchmark design
  • AI

• Physics-Based Force Simulation for Hand-Object Interaction

  We will develop and validate a physics-based pipeline to estimate interaction forces from hand-object motion data. Using an existing dataset that provides synchronized hand and object mesh sequences together with measured interaction forces, we will simulate contact forces from motion and compare them against recorded ground truth.

  more...
  
  • contact: Andela Ilic, Christian Holz
  • semester thesis
  • master's thesis

  • hand-object interaction
  • physics simulation
  • biomechanics
  • contact modeling
  • robotics
  • 3D meshes

• Curved surface routing on discrete surfaces for quad meshing and surface parameterization

  We will develop geometry-processing algorithms for routing coherent curve families on discrete surfaces (triangle meshes) using connection- and field-based transport (in the spirit of trivial connections). The project applies these methods to a classic graphics problem: generating direction fields and separatrix networks that support quad meshing and surface parameterization.

  more...
  
  • contact: Christian Holz
  • master's thesis

  • geometry processing
  • surface parameterization
  • vector fields
  • discrete surfaces

• Practical input detection using an EEG brain-computer interface for Augmented Reality

  We will build and evaluate a practical BCI input method for Augmented Reality using a ready-to-use wearable dry-EEG system. We will create a lightweight Unity MR demo, develop classical and ML-based EEG processing for input and intent detection, and run an experimental study to collect and analyze EEG data under realistic AR conditions.

  more...
  
  • contact: Christian Holz
  • semester thesis
  • master's thesis

  • brain-computer interaction
  • EEG
  • neuroscience
  • Augmented Reality

• Multi-View 3D Hand Pose Estimation from Dynamic Egocentric Captures

  This project explores multi-view 3D hand pose estimation by leveraging several dynamically moving viewpoints to address the persistent challenges of occlusion, viewpoint ambiguity, and fine-grained articulation. The goal is to develop a modern multi-view fusion model that exploits cross-view geometric consistency and spatio-temporal cues to achieve robust hand pose estimation. Beneficial will be our access to a motion simulation engine capable of generating large-scale synthetic multi-camera datasets. This will enable experimentation with synthetic-to-real transfer and state-of-the-art training strategies.

  more...
  
  • contact: Dominik Hollidt
  • semester thesis
  • master's thesis

  • Computer Vision
  • 3D Hand Pose Estimation
  • Multi-Camera
  • Egocentric Vision
  • Multi-View Fusion

Current thesis students

• William Jabbour master's ETH Data Science, D-INFK
• Rafael del Rio Alfonso master's TU/e Electrical Engineering
• Jeremy Bitar master's ETH Computer Science, D-INFK
• Ruizhe Zhu master's ETH Computer Science, D-INFK
• Liuxin Qing master's ETH Electrical Engineering, D-ITET

• Gabriele Mura semester ETH Computer Science, D-INFK

Past Courses and Seminars

• Human-Computer Interaction bachelor course
  • 2025
  • 2024
  • 2023
  • 2022
  • 2021
  • 2020
  • 2019
• Mixed Reality master course
  • 2025
  • 2024
  • 2023
• Mobile Health and Activity Monitoring master course
  • 2025
  • 2024
  • 2023
  • 2022
• Ubiquitous Computing master course
  • 2021
  • 2020
• Adaptive User Interfaces through Machine Learning bachelor seminar
  • 2025
• Egocentric Perception for Human-World Understanding master seminar
  • 2025
• Reinforcement Learning for Understanding and Modeling Human Behavior master seminar
  • 2025
• Advanced Topics in Mixed Reality bachelor seminar
  • 2024
  • 2023
  • 2022
• Advanced Topics in Technical Human-Computer Interaction bachelor seminar
  • 2021
  • 2020