Overview

The DeepHERO Lab operates along two main tracks. The first track addresses both theoretical and applied challenges in deep learning, advancing methodologies to optimize model performance, reliability, and scalability. The second track focuses on heterogeneous computing, exploring the integration and co-design of diverse computational units—such as CPUs, GPUs, FPGAs, DSPs, and AI accelerators—to enhance efficiency and capability in modern AI applications. Together, these tracks enable the lab to push the boundaries of AI and computing for complex, high-performance environments.