Hoi Jun Yoo
Papers
1
Total Citations
4
H-Index
1
About
Hoi Jun Yoo is a prominent researcher specializing in energy-efficient deep neural network (DNN) hardware design, with a particular focus on mobile and embedded AI systems for human-robot interaction (HRI). His work bridges the gap between advanced machine learning algorithms and practical, low-power hardware implementations, addressing one of the most pressing challenges in modern AI deployment: bringing intelligent processing to resource-constrained devices. Among his notable contributions is a variable bit precision DNN processor capable of operating across a 1-bit to 16-bit range, enabling significant energy savings without sacrificing model performance. This work integrates a look-up-table-based processing engine and near-zero skipper architecture to efficiently execute both CNN-based facial emotion recognition and RNN-based emotional dialogue generation within a unified mobile system — a compelling step toward naturalistic human-robot interaction in everyday devices. Yoo's research reflects a deep commitment to co-designing algorithms and hardware to unlock real-world AI applications at the edge. His contributions have garnered recognition within the hardware and AI communities, establishing him as an influential figure in neuromorphic and mobile AI processor design, inspiring students and engineers working at the intersection of machine learning and VLSI systems.
Research Focus
Key Achievements
Top Papers
- 1