Development of High-Speed Indium Phosphide Quantum Dot Neuromorphic Chips for Real-Time Processing in Event-Based

Event-based cameras, or dynamic vision sensors, are transformative in visual data capture, recording scene changes asynchronously with refresh rates often above thousands of events per second. This enables high temporal resolution and efficient data representation, ideal for rapid-response applications. Yet, their immense data output challenges conventional processors.

Neuromorphic computing architectures, modelled on brain-like neural networks, provide an efficient solution by processing asynchronous inputs through spiking neural networks, offering low latency and high parallelism. Indium Phosphide (InP) colloidal quantum dots (QDs) are particularly suited for such systems due to their high electron mobility and direct bandgap, enabling fast neuromorphic devices that can match event-based cameras' extreme speeds.

This CDT project has three key objectives:
1. Develop high-speed neuromorphic devices using InP QDs in electrochemical-gated transistors and memristors through inkjet printing (Hou's lab, PHYSX).
2. Test and validate the chip in Robotics and Autonomous Systems lab to ensure real-time processing of event-based camera data with minimal latency (Ji's lab, ENGIN).
3. Design and implement a seamless high-speed interface between the neuromorphic chip and event-based cameras for continuous, real-time data processing. (Hou's lab and Ji's lab).