Projects

At the Center for Project-Based Learning (PBL), we offer hands-on projects that allow students to apply their knowledge, develop new skills, and collaborate on real-world challenges. Below, you’ll find our current projects open for applications. Whether you’re looking to innovate, prototype, or conduct cutting-edge research, this is your chance to get involved!

Interested in past projects? Click here to explore previous student work and get inspired.

From Events to Graphs: Open-Vocabulary Scene Understanding with Vision-Language Models and Event Cameras

The project involves designing a preprocessing pipeline to adapt event streams for VLM-compatible inputs, adapting scene graph generation to sparse or asynchronous representations, and benchmarking results against conventional RGB-based baselines.

Keywords

Event-Based Vision, Scene Graph Generation, Vision-Language Models, Multimodal Learning

Labels

Semester Project , Bachelor Thesis , Master Thesis , ETH Zurich (ETHZ)

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project...

Published since: 2025-06-19 , Earliest start: 2025-06-09 , Latest end: 2027-01-04

Organization Center for Project-Based Learning D-ITET

Hosts Bonazzi Pietro

Topics Engineering and Technology

Efficient On-Device Foveated Event-Based OCR for Smart Glasses

This project builds upon our existing smart glass platform [2] and explores the development of an ultra-efficient, fully on-device pipeline for text recognition under challenging conditions such as low light and high-speed motion.

Keywords

Computer Vision, Event-Based Vision, Embedded AI, Smart Glasses

Labels

Semester Project , Bachelor Thesis , Master Thesis , ETH Zurich (ETHZ)

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project...

Published since: 2025-06-19 , Earliest start: 2025-06-30 , Latest end: 2026-06-30

Applications limited to ETH Zurich

Organization Center for Project-Based Learning D-ITET

Hosts Bonazzi Pietro

Topics Information, Computing and Communication Sciences , Engineering and Technology

Anchor-Free Eye Tracking on Neuromorphic Event Data

This project investigates anchor-free spiking neural networks (SNNs) for efficient and real-time eye tracking using event-based vision data. Building upon the Retina framework and inspired by the anchor-free detection approach of [2], the goal is to remove anchor mechanisms to simplify training and inference, thus reducing computational overhead and improving generalization.

Keywords

Computer Vision, Spiking Neural Network

Labels

Semester Project , ETH Zurich (ETHZ)

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project...

Published since: 2025-06-19 , Earliest start: 2024-11-24 , Latest end: 2025-11-14

Applications limited to ETH Zurich

Organization Center for Project-Based Learning D-ITET

Hosts Bonazzi Pietro

Topics Information, Computing and Communication Sciences , Engineering and Technology , Behavioural and Cognitive Sciences

Development and Integration of an Ultrawideband Localization System for Tortoise Monitoring in a Zoological Environment

Ultrawideband (UWB) localization is an emerging technology for centimeter-level localization in various applications. This technology enables the development of a flexible and robust system to acquire position data with minimal power requirements. Such systems can be used in various scenarios, one of which is animal tracking in zoological environments, where the system is installed semi-permanently to gather data over an extended time span. This approach allows the collection of valuable data that may shed light on animal behavior, with the ultimate goal of improving animal welfare. By designing an anchor-based UWB localization system, we aim to develop a solution that can be easily installed, automatically localize itself, and upload the collected data via LoRaWAN to the cloud. This reduces power requirements and keeps the system lightweight, robust, and flexible.

Keywords

Ultrawideband (UWB), Localization, Energy Optimized Electronics, LoRaWAN, Animal Tracking

Labels

Master Thesis

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project...

Published since: 2025-06-16 , Earliest start: 2025-08-04

Applications limited to ETH Zurich , Department of Information Technology and Electrical Engineering

Organization Center for Project-Based Learning D-ITET

Hosts Luder Victor

Topics Engineering and Technology

Towards Efficient and Task-General SSL for Event Cameras

Event cameras offer an asynchronous, high-temporal resolution, and sparse data stream that remains underutilized in current self-supervised learning (SSL) approaches.

Keywords

Machine Learning, Computer Vision

Labels

Semester Project , Bachelor Thesis , Master Thesis , ETH Zurich (ETHZ)

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project...

Published since: 2025-06-06 , Earliest start: 2025-06-08 , Latest end: 2026-12-31

Organization Center for Project-Based Learning D-ITET

Hosts Bonazzi Pietro

Topics Engineering and Technology

Ultrasound measurement of microbubble stiffness for in situ detection of protease activity in clinical settings

A new kind of gas-filled microbubble enables the detection of protease activity by using ultrasound imaging techniques. The main goal of this work is to develop a setup that can reliably be used to measure the stiffness of microbubbles, first in a microbubble solution, and then in a model built to simulate the vasculature of a mouse.

Keywords

Ultrasound imaging techniques, microbubbles, signal processing, modeling, analog and digital filter design, circuit design

Labels

Semester Project , Master Thesis

Description

Goal

Contact Details

More information

Open this project...

Published since: 2025-05-28

Applications limited to Department of Information Technology and Electrical Engineering

Organization Center for Project-Based Learning D-ITET

Hosts Villani Federico

Topics Engineering and Technology

Development Of An FPGA-Based Optoacoustic Image Reconstruction Platform for Clinical Applications

Optoacoustic (OA) imaging is a hybrid imaging method that enables deep tissue imaging with a high spatial resolution by combining optical illumination with ultrasound detection. The goal of this student project is to devise a parallel HW accelerator and explore different HLS code optimizations to achieve the best performance for OA image reconstruction on an FPGA in real-time.

Keywords

Digital design (HDL), biomedical imaging, image reconstruction, hardware acceleration, Linux

Labels

Semester Project , Bachelor Thesis , Master Thesis

Description

Optoacoustic (OA) imaging is a hybrid imaging method that enables deep tissue imaging with a high spatial resolution by combining optical illumination with ultrasound detection. OA imaging is based on the OA effect, where the ultrasound (US) waves are generated when incident laser energy is absorbed by different molecules leading to local temperature rise. The acoustic pressure propagating from the molecules is acquired using various ultrasonic transducers. Finally, with the help of different reconstruction algorithms, the initial pressure distribution from the acquired pressure waves is recovered. In practice, most of these reconstructions are carried out on workstations making OA imaging systems bulkier and more complex. Hence realizing OA image reconstruction on FPGAs will enhance the adaptability of OA imaging systems with point-of-care (bedside) applications due to their small size and high computation power. With our MATLAB and C++-based delay and sum-based image reconstruction algorithms, we can successfully reconstruct OA images from pre-acquired raw OA signals. We are currently working on the implementation of FPGA based delay and sum-based image reconstruction algorithm with pre-acquired raw OA signals using high-level synthesis (HLS). Your task includes devising a parallel HW accelerator and exploring different HLS code optimizations to achieve the best performance. The algorithms will be tested on an existing FZ3 Card, a powerful deep-learning accelerator card based on Xilinx Zynq UltraScale+ ZU3EG MPSoC. The overall goal will be to streamline the raw OA data and perform accelerated image reconstruction on the FPGA in real-time. **Prerequisites** - FPGA Programming (VLSI 1 course or similar) - C++ Programming - MATLAB or Python for evaluation of results - Image processing algorithms (beneficial) - Knowledge of Linux (beneficial) **Type of work** - 10% Literature study - 20% Algorithm evaluation - 30% Firmware design and optimization - 30% Measurements and validation - 10% Documentation

Goal

Contact Details

More information

Open this project...

Published since: 2025-05-28

Applications limited to ETH Zurich , Department of Information Technology and Electrical Engineering

Organization Center for Project-Based Learning D-ITET

Hosts Villani Federico

Topics Engineering and Technology

Development Of A Test Bed For Ultrasonic Transducer Characterization (1 S/B)

The main goal of this work is to develop a modular system for the characterization and tuning of ultrasonic transducers both in hard- and software Due to the intended modularity of the system in soft- and hardware, we can guarantee a high flexibility of the setup. This means that the system can be adapted in operation for a wide variety of transducer types and setups.

Keywords

Ultrasound, Flexible transducers, Mixed signal design, Network theory, Simulation

Labels

Semester Project , Bachelor Thesis

Description

Goal

Contact Details

More information

Open this project...

Published since: 2025-05-28

Applications limited to Department of Information Technology and Electrical Engineering

Organization Center for Project-Based Learning D-ITET

Hosts Villani Federico

Topics Engineering and Technology

A FPGA-based data streaming system that enables real-time monitoring of cell culture and neuroactivities

This project aims to create a FPGA-based interface capable of streaming live data from a ETH-developed CMOS biosensor and monitor neuroactivity and cell cultures in real time

Keywords

Sensing, ADC, FPGA, High-performance computing, biosensing

Labels

Semester Project , Master Thesis

Description

**Introduction** The physiology of living cells and tissues is of large interest to pharmaceutical industry for new drug development, as important information such as the drug efficacy and toxicity can be readily assessed in vitro. As such, the development of CMOS biosensors for cell monitoring and manipulation is essential. Many existing CMOS cellular sensors, however, suffer from one major limitation: of the many pixels on the sensing platform, only a handful of channels can be read concurrently, limiting our understanding of the cell culture as a whole. A newly developed platform achieves 20k concurrent channel reading, at 20kHz frame rate and 10 bit resolution. However, the vast output data stream that needs to be accessed and monitored calls for a high-speed FPGA/GPU based system with a compact, and incubator-compatible hardware design. **Hardware design** Central to this project is the utilization of a powerful, state of the art Xilinx FPGA. The choice of the FPGA is primarily limited by the choice of interface and by the power consumption due to the lack of active dissipation. Furthermore, depending on the FPGA choice and the student’s interest, the scope of the thesis / project may be either a full custom hardware built upon the FPGA, or a simpler board with mounted system-on-module. **Prerequisites** - Familiarity with FPGA programming (VLSI 1 or similar) - Programming experience in Python and C++ - PCB Design (Helpful but not required) **Possible Project Phases** 1. Proof-of-concept FPGA system: - Reads data from some of the ADC outputs (even just 1 or 2) into the FPGA - Streams dummy data at a speed near to 10 Gbit to a storage device - Streams the ADC output to the PC, and visualizes a selectable subset of data 2. Hardware design: - Developed based on the proof-of-concept system, the hardware system reads from all ADC channels and stream the entire data to a storage device. 3. Hardware design validation: - The hardware design is able to stream at least some ADC outputs to the PC 4. Data compression algorithm design: - The FPGA implements an algorithm that leverages large on-FPGA memory to explore possibilities for data compression

Goal

Contact Details

More information

Open this project...

Published since: 2025-05-28

Applications limited to Department of Information Technology and Electrical Engineering

Organization Center for Project-Based Learning D-ITET

Hosts Villani Federico

Topics Engineering and Technology

Embedded Firmware development for high performance PTU

Join an exciting opportunity to develop and extend low-level firmware for a high-performance pan-tilt unit (PTU) powered by an STM32 microcontroller. This open-source project focuses on enhancing existing motor control firmware with modern networking and remote-control capabilities to create a modular, low-latency control system. Ideal for students looking to make meaningful contributions to embedded systems in robotics or automation.

Keywords

ptu,embedded,c,firmware,low-level,tcp,udp,protocol

Labels

Semester Project , Master Thesis , ETH Zurich (ETHZ)

Description

Goal

More information

Open this project...

Published since: 2025-05-20 , Earliest start: 2025-05-20 , Latest end: 2026-03-31

Applications limited to ETH Zurich

Organization Center for Project-Based Learning D-ITET

Hosts Mandula Jakub

Topics Information, Computing and Communication Sciences , Engineering and Technology

Text2Video: Fast Text-Driven Video Generation

This project aims to leverage existing text-to-video generative models [1, 2] to produce video content based on textual prompts. Unlike conventional approaches that prioritize high quality, the main objective of this project is low-latency efficient video generation

Keywords

Video Generation, Efficient Inference, Real-Time Synthesis

Labels

Semester Project , Bachelor Thesis , Master Thesis , ETH Zurich (ETHZ)

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project...

Published since: 2025-05-14 , Earliest start: 2025-06-02 , Latest end: 2026-10-05

Organization Center for Project-Based Learning D-ITET

Hosts Bonazzi Pietro

Topics Information, Computing and Communication Sciences , Engineering and Technology

Low Power Wireless On-Body Communication for Body-attached Sensor Nodes

Wearable, wirelessly connected sensors have become a common part of daily life, evolving step by step from their roots in sports and fitness to play a pivotal role in shaping the future of personalized healthcare. A key challenge in this evolution is designing devices that are unobtrusive, highly integrated, and energy efficient. These design requirements inherently demand smaller batteries, which must also support the significant power consumption of wireless communication interfaces. Capacitive Human Body Communication (HBC) offers a promising, power-efficient alternative to traditional RF-based communication, enabling point-to-multipoint data and energy exchange. By using the conductive properties of the human body, a privacy-preserving wireless personal body area network (WBAN) can be created. Several low-power sensors such as ECG-tracker and insulin pumps can act as leaf devices, sending personal data to a body-central gateway, such as a smartwatch that further processes the data and establishes a connection to the cloud. .

Keywords

Hardware design, Firmware design, analog circuit design, digital circuit design, signal processing, field evaluation

Labels

Energy Harvesting (PBL) , Semester Project , Master Thesis , Microcontroller (PBL) , PCB Design (PBL) , Software (PBL) , Firmware (PBL) , Machine Learning (PBL) , Biomedical (PBL) , Wearables (PBL)

Description

Goal

Contact Details

More information

Open this project...

Published since: 2025-05-12

Applications limited to Department of Information Technology and Electrical Engineering

Organization Center for Project-Based Learning D-ITET

Hosts Schulthess Lukas

Topics Engineering and Technology

Pushing Ultrasound to the Limit: 100 Gbit Ethernet Interface for Imaging at the Edge of Physics

Next-generation ultrasound (US) imaging demands exceptionally high data throughput, exceeding 90 Gbps for ultrafast applications. This project harnesses datacenter networking techniques—100G Ethernet, RDMA, and HPC-grade storage—to enable real-time streaming of raw US data at minimal latency. By integrating advanced optical interfaces and bypassing conventional CPU-intensive workflows, this project aims to achieve sustained multi-gigabit performance, paving the way for cutting-edge imaging analytics and machine learning within modern datacenter environments

Keywords

Datacenter networking, 100G Ethernet, HPC, RDMA, Infiniband, high-speed streaming, ultrasound, Linux kernel programming

Labels

Semester Project , Master Thesis , FPGA (PBL) , Software (PBL) , Firmware (PBL) , Biomedical (PBL)

Description

Goal

Contact Details

More information

Open this project...

Published since: 2025-04-08 , Earliest start: 2025-07-01

Applications limited to Department of Computer Science , Department of Information Technology and Electrical Engineering

Organization Center for Project-Based Learning D-ITET

Hosts Villani Federico

Topics Engineering and Technology

Transformer-based model for Gesture Recognition

A major shortcoming of current gesture recognition methods is lack of robustness w.r.t. wearing position, changing users, background activities, etc. Addressing this requires research in various areas, such as new network architectures, model generalization methods, data-collection and data-engineering methods, and the specific choice of gestures. The project aims to explore some or all of these points.

Keywords

Times-Series, Gesture Recognition

Labels

Semester Project , Master Thesis , ETH Zurich (ETHZ)

PLEASE LOG IN TO SEE DESCRIPTION

More information

Open this project...

Published since: 2025-02-10 , Earliest start: 2025-02-17 , Latest end: 2025-12-15

Applications limited to ETH Zurich , EPFL - Ecole Polytechnique Fédérale de Lausanne

Organization Center for Project-Based Learning D-ITET

Hosts Bonazzi Pietro

Topics Engineering and Technology