Research
Ongoing Research Projects
Custom Hardware and Embedded/Cyber-Physical System Design for Privacy-Preserving Applications
FPGA- and heterogeneous computing platform-based hardware accelerators for fully homomorphic encryption primitives
Homomorphic encryption-based real-world applications involving logical and non-polynomial functions
Sponsors: RIT FEAD (2024), NSF CRII (2021-2023, 2 years), IITP (2020-2023, 4 years), UW RRF (2020-2021, 1 year)
Publications: [ReConFig'19], [FCCM'20], [RTCSA'21], [GLSVLSI'22], [MMSP'22-1], [MMSP'22-2], [ISCAS'23-1], and [GLSVLSI'24]
Collaborators: Seoul National University and Intel
Custom Hardware Design for Approximate Arithmetic
FPGA-based hardware accelerators for approximate operations on real numbers
Publications: [GLSVLSI'19], [ISCAS'21], [ISCAS'23-2], [GLSVLSI'23], and [TVLSI'24]
Past Research Projects
Custom Hardware Design for Machine Learning Problems
FPGA-based accelerators of binary convolutional neural networks
Publication: [FCCM'18]
Management on Non-Volatile Memory Systems
Hot address-based wear-leveling for phase-change memory
Sponsor: SK Hynix (2016-2017)
Publications: [CAL'19] and two patents
Fast and Energy-Efficient RGBW Display Systems
Look-up table and interpolation-based RGB-to-RGBW conversion
Publications: [APCCAS'16] and [TCAS-II’18]
Custom Hardware Design for Image/Video Processing
Hardware accelerators for frame memory compression
Sponsors: LG Display (2012-2015), ETRI (2015-2016)
Publications: [Displays'15], [TMM'16], [TCE'16], [JDT'16], [ISCAS'16], [JETCAS'16], [JSTS'17-1], and [JSTS'17-2]
Cost-Effective Computer Vision Algorithms
Gesture recognition-based virtual keyboard system
Trajectory detection using pedestrian and surrounding information
Face detection and tracking using H.264/AVC information
Publications: [ICEIC'12], [SPC'16], and [THMS'22]
Acknowledgement
Grateful to the National Science Foundation (NSF), Institute for Information and Communications Technology Promotion (IITP), University of Washington Royalty Research Fund (RRF) program, Rochester Institute of Technology Faculty Education and Development (FEAD) program for the funds
Grateful to Intel and AMD-Xilinx for their FPGA donations
