Papers
3
Total Citations
36
H-Index
3
About
Seungjin Lee is a leading researcher in energy-efficient artificial intelligence hardware, specializing in embedded neural network accelerators and bio-inspired vision processors for mobile and robotic platforms. His work focuses on bridging the gap between high-performance AI algorithms and the stringent power constraints of portable devices, enabling real-time intelligent functions like object detection and human-computer interaction. Lee’s most impactful contribution is the development of a 57mW mixed-mode neuro-fuzzy accelerator for multi-core processors, which demonstrated that complex AI tasks could be executed with minimal energy consumption, garnering 18 citations. He also pioneered a 201.4 GOPS real-time multi-object recognition chip that leveraged a bio-inspired neural perception engine, achieving 15 citations for its ability to process multiple objects simultaneously—a significant leap over prior single-object systems. Additionally, his 81.6 GOPS vision processor for mobile robots, based on the Scale Invariant Feature Transform (SIFT) algorithm, optimized task and data-level parallelism to deliver power-efficient object recognition. Lee’s work has been instrumental in advancing intelligent embedded systems, with his chips achieving remarkable performance-per-watt ratios (e.g., 496mW for 201.4 GOPS), making him a key figure in the evolution of low-power AI hardware for real-world applications.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3