Personal profile
Research Focus
My research interests are mainly in resilient and real-time ML and AI accelerators, especially in how hardware-efficient methods can be used to make machine learning systems more practical on FPGA and other resource-constrained platforms. At the moment, I am particularly interested in approximate computing, reduced-precision arithmetic, and the use of alternative number formats such as POSIT. What I am trying to understand is whether these methods can reduce hardware cost and improve efficiency without causing too much loss in numerical quality or reliability.