Developing real-time low-power multi-channel speech separation algorithms
Unsupervised error detection using a spiking neural network - Semester project
New projects from devices to systems with the Sensors group
Designing real time interaction between machine and birds
Deep Learning for Time-Dependent Behavioural Data – Semester Project
Data Compression – Paid Internship
Exploration of Hypothetical Astrocyte Functions – Semester Project
Model-Based Estimators of Functional Connectivity for Calcium Imaging Data – Masters Project
Bridging Function and Anatomy in Wide-field Calcium Imaging Data - Masters Project
Neuromorphological changes during vocal development in zebra finches - Investigating how learning to sing shapes a songbird’s brain
Neuronal population dynamics in a learned motor behavior.
Investigate the neural code along the output axis of the brains memory system
Neuromorphic chip interface with a microcontroller - Master project
This project will allow to explore input/output solutions of spikes generated by (or sent to) neuromorphic chips using the Teensy 4.0 platform (ARM Cortex M7), and to characterize temporal (1us resolution) and bandwidth requirements.
Design and implementation of dissolvable elastic electrodes for stable neuronal activity recording
High-speed tracking of head pose and facial expression using dynamic vision sensor event cameras
Detecting iEEG High Frequency Oscillations Using Event-mapped Hyperdimensional Computing
Biophysical characterization of the subdural local field potentials in mouse neocortex and fabricating low-impedance flexible electrode arrays
Fabrication of flexible shank-electrodes for long-term multi-area electrophysiological recordings in the prefrontal areas of behaving rats
Reinforcement Learning of Visual Decision Making in RNNs (Master Thesis)
Development of Electronic Hardware and Control Software for a Miniaturized Microscope
Text Embeddings Optimized for Distance Computations (MSc Thesis)
The goal of this project is to develop new methods for representation learning of documents and sentences, that are trained to approximate the distance between two documents.
Intrinsic Optical Imaging for the Mapping of Cortical Responses
Focused Ultrasound mediated Drug delivery: In-Vitro Tests
Online Monitoring of Vital Signs & Anesthesia during Neural Recordings
Classification of Radio Signals on a neuromorphic processor in Space
Fully automated extraction of neural signals from imaging data
Machine learning for neuronal activity
Controlling sequential movements with neural networks in neuromorphic hardware
Learning experience maps on a neuromorphic chip with Khepera robot equipped with a dynamic vision sensor
Neural network implementation in neuromorphic hardware for unsupervised learning of MNIST digits
Computer vision based reconstruction of neuromorphological features in the songbird’s syrinx on 3D tomograms (PSI)
Syrinx's biomechanics in songbirds with in vivo high-speed 2D tomograms (PSI)
Reinforcement learning of human vocal behavior
We study reinforcement learning of fundamental frequency (pitch) in songbirds and humans. When birds receive aversive reinforcement for low-pitch syllables they successfully learn to increase the syllables’ pitch.
Analysis of birdsong development and automated clustering of song syllables
During early development, young songbirds such as the Zebra Finch learn acoustically complex but stereotyped sequential behaviors which are termed "songs". Furthermore, zebra finches learn only one song in their lifetime, making the problem of developmental song analysis tractable.
Psychophysical Theory of Human Pitch Processing
We study the mechanisms of fundamental frequency (pitch) adaptation of songbird and human vocalizations. Adaptation can be induced as a response to distortions of pitch feedback.
Neuronal controllers for cognitive robots
In our group “Neuromorphic Cognitive Systems", we develop neuronal architectures that allow robots to generate behavior (e.g. navigate in an environment, avoid obstacles, pursue targets), to form memories (e.g. build a map), and learn.
Real-time feedback-controlled delivery of neuromodulators for non-invasive high-resolution modulation of brain
We are employing MRI guided focused ultrasound based delivery of neuromodulators to control activity of specific brain micro circuits and subsequent cognitive behavior, which has both fundamental and medical applications.
Deep neural networks for auditory scene analysis
A Spike based ADC for calcium bio-signal recording in the mouse brain
Depth vision with a neuromorphic sensor on a robotic vehicle
Neuromorphic controller for an autonomous quadcopter
UZH - ETH Zurich students:
If you are interested in working on one of the projects listed above, there is always the possibility of defining sub-projects and objectives suitable for Master theses and Semesterarbeits. Please contact your prospective supervisor directly, and arrange a meeting with him or her.
Unfortunately we do not have any means of providing financial support for internships or thesis projects. However, if you find alternative sources of funding or can provide your own financial support, we can define student/thesis projects, provide the necessary lab and office space, and supervise your work. Given the experimental nature of the work being carried out in our lab, involving the design or use of custom analog VLSI devices, it only makes sense to consider project durations of at least 6 months.
Please make sure you can demonstrate your ability to financially support yourself (mainly for visa reasons) before contacting your prospective supervisor.