Papers

2

Total Citations

75

H-Index

2

About

Sanghoon Kang is a leading researcher in energy-efficient deep learning hardware, specializing in mobile and real-time AI systems. His work focuses on accelerating deep neural networks for autonomous platforms, particularly in reinforcement learning and 3D perception. Kang’s most notable contribution is the development of a 2.1 TFLOPS/W mobile deep reinforcement learning accelerator, featuring a transposable processing element array and experience compression—a design that achieves exceptional energy efficiency for real-time action control in robots and autonomous systems. This work, published in 2019, has garnered 59 citations, highlighting its impact on low-power AI hardware. More recently, Kang introduced the DSPU, a 281.6 mW real-time depth signal processing unit for dense RGB-D data acquisition and 3D bounding box extraction, enabling mobile platforms like AR devices and autonomous robots to operate at over 30 fps with minimal power consumption. His innovations bridge the gap between high-performance deep learning and the stringent power constraints of edge devices, advancing practical AI deployment in robotics and augmented reality.

Research Focus

Key Achievements

2
H-Index
2
Papers
75
Total Citations
38
Avg Citations/Paper
🏆 Most Cited Paper
A 2.1TFLOPS/W Mobile Deep RL Accelerator with Transposable PE Array and Experience Compression
59 citations · 2019
📈 Most Prolific Year: 2019 (1 Papers)
🤝 Key Collaborators: 11
🏛 Institutions: Korea Advanced Institute of Science and Technology

Top Papers

  1. 1
  2. 2

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago