Semester and Thesis Projects

The ETH AI Center offers a wide range of semester and thesis projects for students at ETH Zurich, as well as other universities. Please see the list below for projects that are currently available.

How do you publish a thesis or semester project with the ETH AI Center?

  • Academia: If you are affiliated with the ETH AI Center as faculty member, post- or doctoral fellow, please add the following affiliation to your Sirop account. If you tag your thesis project with this affiliation, it should appear in the list below.
    • 'ETH Competence Center - ETH AI Center (ETHZ)'
  • Industry: If you represent a company that has a corporate partnership with the ETH AI Center, please contact .

Need help?

ETH Zurich uses SiROP to publish and search scientific projects. For more information visit sirop.org.

Development of a Web Application for Evaluating and Actively Adapting Personalized ASR Models

Automatic speech recognition (ASR) is key to human-computer interaction, but struggles with the variability of children's speech, especially in cases of speech impairment. Our group developed a personalized ASR approach for such children, even with limited data. To test and refine this system using active learning, we aim to build a web app that lets users evaluate their model, upload new speech samples, and trigger fine-tuning based on transcription errors.

Keywords

Non-normative speech recognition, personalized ASR, web application, human-computer interaction, low-resource adaptation, active learning, user feedback, model fine-tuning

Labels

Semester Project

Contact Details

More information

Open this project... 

Published since: 2025-07-27 , Earliest start: 2025-07-18

Organization ETH Competence Center - ETH AI Center

Hosts Moure Pehuen , Böhringer Roman

Topics Information, Computing and Communication Sciences

Master Thesis Opportunities in Agentic AI Applications at the ETH Agentic Systems Lab

Join the newly founded Agentic Systems Lab at ETH Zurich to develop AI systems with real-world impact. We are inviting 10–20 Master’s students to join our “Genesis Batch 2025” this semester (until the end of the year). Choose from three tailored tracks: Explorer (startup/industry collaboration), Builder (AI for education, science & healthcare), and Bring Your Own Project (develop your own AI venture). Collaborations with leading startups such as Browser-Use (San Francisco) and nunu.ai (Zurich) are already underway, with more to follow.

Keywords

AI, agentic systems, master thesis, ETH Zurich, startups, applied AI, AI ventures, research, healthcare AI, education AI, scientific applications, prototyping, automation

Labels

Master Thesis , ETH Zurich (ETHZ)

Description

Goal

Contact Details

More information

Open this project... 

Published since: 2025-07-25 , Earliest start: 2025-07-28 , Latest end: 2026-03-31

Organization ETH Competence Center - ETH AI Center

Hosts Da Conceição Barata Filipe

Topics Engineering and Technology , Education

RA position: CoMind: Joint Interaction Understanding through Intent Prediction

Understanding human intent and anticipating future actions is crucial for enabling seamless human-robot interaction in real-world tasks. Intention prediction not only facilitates smoother collaboration with robots but also represents a fundamental challenge in the development of intelligent systems. In this project, we aim to collect a dedicated interaction dataset using the Meta Aria glasses and use the collected data to develop multimodal learning models that are capable of predicting human intent and anticipating future behaviors across a variety of tasks.

Keywords

egocentric vision, intent prediction, anticipation, multimodal learning

Labels

Semester Project , Master Thesis , Student Assistant / HiWi , ETH Zurich (ETHZ)

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2025-07-14 , Earliest start: 2025-07-14

Organization Computer Vision and Geometry Group

Hosts Wang Xi , Kaufmann Manuel , Chen Jiaqi , Gavryushin Alexey

Topics Information, Computing and Communication Sciences

ZipMPC for Vision-Based Autonomous Racing

This project aims to extend the ZipMPC framework, a method that enables computationally efficient long-horizon model predictive control (MPC) using a compressed, context-aware cost function combined with short-horizon MPC, by incorporating vision-based context. Specifically, we propose to learn the context pertaining to reference trajectory from images, enabling long-horizon vision-informed control, while maintaining the efficiency and constraint satisfaction of short-horizon MPC.

Keywords

Model predictive control, vision-based control, imitation learning, context-dependent control

Labels

Master Thesis

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2025-07-08

Organization ETH Competence Center - ETH AI Center

Hosts Trisovic Jelena

Topics Information, Computing and Communication Sciences , Engineering and Technology

Brain Machine Interface - Visual Neuroprosthetics

Join the Sensors Group at the Institute of Neuroinformatics (INI) to develop next-generation visual neuroprosthetics and advance the future of brain-machine interfaces! Topics Include: - develop neural networks that learn optimal stimulation patterns - utilize recurrent/spiking neural networks for creating stimulation patterns - implementing real-time computation on embedded platforms (FPGA, uC, jetson) - investigating closed-loop control strategies for electrical brain-stimulation Application process: Write us about your interests, include CV and transcript, and we can arrange a meeting. We can supervise students from UZH and ETH. We offer semester projects as well as bachelor's and master's theses projects.

Keywords

brain machine interface, visual neuroprosthetics, bmi, neural networks, Real-time computation, Embedded platforms, FPGA, Closed-loop control, neural recording analysis, Control systems, Deep learning, Verilog, Vivado, hls4ml, Hardware acceleration, Jetson, VR, Android, Unity/Blender

Labels

Semester Project , Master Thesis , ETH Zurich (ETHZ)

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2025-06-17 , Earliest start: 2025-02-01 , Latest end: 2026-12-31

Applications limited to ETH Zurich , University of Zurich

Organization ETH Competence Center - ETH AI Center

Hosts Moure Pehuen , Liu Shih-Chii , Hahn Niklas

Topics Information, Computing and Communication Sciences , Engineering and Technology

Learning a Simulation-Trained Safety Critic for Safe Online Learning in Legged Robots

This project focuses on developing a safety critic—a model that predicts the safety of robot states—to enable safe online learning on legged robotic hardware. The safety critic is trained in simulation using labeled data from diverse robot behaviors, identifying states likely to lead to failure (e.g., falls). Once trained, the critic is deployed alongside a learning policy to restrict unsafe exploration, either by filtering dangerous actions or shaping the reward function. The goal is to allow adaptive behavior on real hardware while minimizing physical risk.

Keywords

safety critic, online learning

Labels

Master Thesis

Description

Contact Details

More information

Open this project... 

Published since: 2025-06-16

Organization Robotic Systems Lab

Hosts Li Chenhao , Li Chenhao , Li Chenhao , Li Chenhao

Topics Engineering and Technology

Active System Identification for Efficient Online Adaptation

This project proposes a novel single-stage training framework for system identification in legged locomotion, addressing limitations in the conventional two-stage teacher-student paradigm. Traditionally, a privileged teacher policy is first trained with full information, followed by a student policy that learns to mimic the teacher using only state-action histories—resulting in suboptimal exploration and limited adaptability. In contrast, our method directly trains a policy to regress privileged information embeddings from its history while simultaneously optimizing for an active exploration objective. This objective is based on maximizing mutual information between the policy’s state-action trajectories and the privileged latent variables, encouraging exploration of diverse dynamics and enhancing online adaptability. The approach is expected to improve sample efficiency and robustness in deployment environments with variable dynamics.

Keywords

Active Exploration, System Identification, Online Adaptation

Labels

Master Thesis

Description

Contact Details

More information

Open this project... 

Published since: 2025-06-06

Organization Robotic Systems Lab

Hosts Li Chenhao , Li Chenhao , Li Chenhao , Li Chenhao

Topics Engineering and Technology

Learning from Online Demonstrations via Video Diffusion for Local Navigation

This project introduces a framework for local navigation skill acquisition through online learning from demonstrations, bypassing the need for offline expert trajectories. Instead of relying on pre-collected data, we use video diffusion models conditioned on semantic text prompts to generate synthetic demonstration videos in real time. These generated sequences serve as reference behaviors, and the agent learns to imitate them via an image-space reward function. The navigation policy is built atop a low-level locomotion controller and targets deployment on legged platforms such as humanoids and quadrupeds. This approach enables semantically guided, vision-based navigation learning with minimal human supervision and strong generalization to diverse environments.

Keywords

Learning from Demonstrations, Video Diffusion, Semantic Conditioning

Labels

Master Thesis

Description

Contact Details

More information

Open this project... 

Published since: 2025-06-06

Organization ETH Competence Center - ETH AI Center

Hosts Li Chenhao , Li Chenhao , Li Chenhao , Li Chenhao

Topics Information, Computing and Communication Sciences

Learning World Models for Legged Locomotion (Structured legged world model)

Model-based reinforcement learning learns a world model from which an optimal control policy can be extracted. Understanding and predicting the forward dynamics of legged systems is crucial for effective control and planning. Forward dynamics involve predicting the next state of the robot given its current state and the applied actions. While traditional physics-based models can provide a baseline understanding, they often struggle with the complexities and non-linearities inherent in real-world scenarios, particularly due to the varying contact patterns of the robot's feet with the ground. The project aims to develop and evaluate neural network-based models for predicting the dynamics of legged environments, focusing on accounting for varying contact patterns and non-linearities. This involves collecting and preprocessing data from various simulation environment experiments, designing neural network architectures that incorporate necessary structures, and exploring hybrid models that combine physics-based predictions with neural network corrections. The models will be trained and evaluated on prediction autoregressive accuracy, with an emphasis on robustness and generalization capabilities across different noise perturbations. By the end of the project, the goal is to achieve an accurate, robust, and generalizable predictive model for the forward dynamics of legged systems.

Keywords

forward dynamics, non-smooth dynamics, neural networks, model-based reinforcement learning

Labels

Master Thesis

Description

Contact Details

More information

Open this project... 

Published since: 2025-06-06

Organization Robotic Systems Lab

Hosts Li Chenhao , Li Chenhao , Li Chenhao , Li Chenhao

Topics Engineering and Technology

Exploring Diffusion-Based Language Models for Medical Report Generation

Recent advances in Natural Language Processing (NLP) have been driven by large language models (LLMs) such as ChatGPT, DeepSeek, and LLaMA. These models are autoregressive by design, generating text word-by-word based on a fixed sequence. While powerful, this approach can lead to limitations, especially when precision and factual consistency are essential. A new class of models — diffusion-based language models — is emerging as a promising alternative. Unlike autoregressive models, these generate and iteratively refine full text sequences, potentially enabling a more robust and coherent generation process. However, their behavior and performance in real-world applications are still largely unexplored. Are you excited about contributing to cutting-edge NLP research, curious about how these models might perform in sensitive, high-precision domains like medicine — and interested in working on a thesis that bridges academic research and real-world startup innovation? If so, then this project is for you. We offer the opportunity to investigate the potential of this new generation of LLMs for reliable summarization, in the context of an applied industry setting.

Keywords

NLP, Large Language Models, Diffusion Models, Medical Text Generation, Deep Learning, Transformer Architectures, Model Evaluation, Thesis Collaboration

Labels

Master Thesis

Description

Goal

Contact Details

More information

Open this project... 

Published since: 2025-06-04 , Earliest start: 2025-06-15 , Latest end: 2026-06-01

Organization ETH Competence Center - ETH AI Center

Hosts Ash Elliott

Topics Medical and Health Sciences , Information, Computing and Communication Sciences

Bridging the Gap: Enabling Soft Actor-Critic for High-Performance Legged Locomotion

Proximal Policy Optimization (PPO) has become the de facto standard for training legged robots, thanks to its robustness and scalability in massively parallel simulation environments like IsaacLab. However, alternative algorithms such as Soft Actor-Critic (SAC), while sample-efficient and theoretically appealing due to entropy maximization, have not matched PPO’s empirical success in this domain. This project aims to close that performance gap by developing and evaluating modifications to SAC that improve its stability, scalability, and sim-to-real transferability on legged locomotion tasks. We benchmark SAC against PPO using standardized pipelines and deploy learned policies on real-world quadruped hardware, pushing toward more flexible and efficient reinforcement learning solutions for legged robotics.

Keywords

Legged locomotion, Soft Actor-Critic, Reinforcement learning, Sim-to-real transfer

Labels

Master Thesis

Description

Contact Details

More information

Open this project... 

Published since: 2025-05-30

Organization ETH Competence Center - ETH AI Center

Hosts Li Chenhao , Li Chenhao , Li Chenhao , Li Chenhao

Topics Information, Computing and Communication Sciences

Handling Long-Tail Distributions in Dense Regression for Forest Biomass and Tree Height Estimation

This master thesis project investigates strategies to address long-tail distributions in dense regression tasks, with a focus on forest-related remote sensing applications. The underrepresentation of high-value targets often degrades model performance in critical areas. The student will explore both learning-based and sampling-based approaches to mitigate this issue. Methods will be evaluated on two benchmark tasks using pre-processed datasets. The goal is to develop and assess robust techniques that improve predictive accuracy in imbalanced scenarios, with implications extending beyond forest monitoring to other domains facing similar distribution challenges.

Keywords

Machine Learning, Computer Vision, Imbalanced Learning, Long tail, Environment

Labels

Master Thesis

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2025-05-26 , Earliest start: 2025-07-01 , Latest end: 2026-02-28

Applications limited to ETH Zurich , University of Zurich

Organization ETH Competence Center - ETH AI Center

Hosts Sialelli Ghjulia

Topics Information, Computing and Communication Sciences

Learning Terrain Traversal from Human Strategies for Agile Robotics

Teaching robots to walk on complex and challenging terrains, such as rocky paths, uneven ground, or cluttered environments, remains a fundamental challenge in robotics and autonomous navigation. Traditional approaches rely on handcrafted rules, terrain classification, or reinforcement learning, but they often struggle with generalization to real-world, unstructured environments.

Keywords

3D reconstruction, egocentric video, SMPL representation

Labels

Semester Project , Master Thesis

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project... 

Published since: 2025-05-21 , Earliest start: 2025-05-26

Organization Computer Vision and Geometry Group

Hosts Wang Xi , Frey Jonas , Patel Manthan , Kaufmann Manuel , Li Chenhao

Topics Information, Computing and Communication Sciences

JavaScript has been disabled in your browser