My current publications are mostly related to approximate computing, unary computing and reconfigurable computer architecture, while my research covers approximate computing, unary computing, quantum computing, reconfigurable computer architecture, and security in deep learning systems.


RAVEN stands for a Reconfigurable Architecture for Varying Emerging Neural Networks. It is designed to be compatible for more complicated neural networks in the future, propelled by approximate computing and network theory.


Unary computing utilizes serial bit streams as data representation, with which both the system latency and the hardware cost can be reduced. Exploring the efficient unary computing kernel and automating the deployment is appealing for real world appliciations.