Integrated Circuit Design for Event-based Vision Sensors

Supervisors:

Primary supervisor Prof André van Schaik

Description:

Event-based vision sensors are the most popular neuromorphic sensor currently, and used for numerous projects at ICNS and elsewhere. However, the current crop of sensors generates its events in a very simple way: by detecting a temporal change in contrast independently at each pixel. While this has proven very useful for some applications, there are other applications where a different method to generate the events would be preferable. This project aims to design such methods and the sensor chips to implement them. This will involve both low-power analogue circuit and digital circuit design.

Eligibility criteria:

Expertise in Matlab, Python, or C++ for software simulation. Basic experience in IC design.