AISys Lab. ECE, Seoul National University
member portrait Kanghyun Choi PhD Student
Email

GitHub

Google Scholar

CV

Search for Kanghyun Choi's papers on the Publications page

I am a graduate student in the Department of Electrical and Computer Engineering at Seoul National University. My research interests are mainly focusing on compressing heavy neural network architecture by neural network quantization.

Publications

DANCE++: Differentiable Accelerator/Network Co-Exploration with Hard Constraints and Data-Free Training for Real-World Scenarios
DANCE++: Differentiable Accelerator/Network Co-Exploration with Hard Constraints and Data-Free Training for Real-World Scenarios
Kanghyun Choi, Deokki Hong, Hyeyoon Lee, Joonsang Yu, Noseong Park, Youngsok Kim, Jinho Lee
IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD)   ·   Jun 2025   ·   To appear
SCI AI/ML Arch.
MimiQ: Low-Bit Data-Free Quantization of Vision Transformers with Encouraging Inter-Head Attention Similarity
MimiQ: Low-Bit Data-Free Quantization of Vision Transformers with Encouraging Inter-Head Attention Similarity
Kanghyun Choi, Hyeyoon Lee, Dain Kwon, SunJong Park, Kyuyeun Kim, Noseong Park, Jonghyun Choi, Jinho Lee
AAAI Conference on Artificial Intelligence (AAAI)   ·   Feb 2025   ·   10.1609/aaai.v39i15.33761
Top Conf. AI/ML
DataFreeShield: Defending Adversarial Attacks without Training Data
DataFreeShield: Defending Adversarial Attacks without Training Data
Hyeyoon Lee, Kanghyun Choi, Dain Kwon, SunJong Park, Mayoore Selvarasa Jaiswal, Noseong Park, Jonghyun Choi, Jinho Lee
International Conference on Machine Learning (ICML)   ·   May 2024   ·   10.48550/arXiv.2406.15635
Top Conf. AI/ML
DANCE: Differentiable Accelerator/Network Co-Exploration
DANCE: Differentiable Accelerator/Network Co-Exploration
Kanghyun Choi1, Deokki Hong1, Hojae Yoon1, Joonsang Yu, Youngsok Kim, Jinho Lee
Design Automation Conference (DAC)   ·   Dec 2021   ·   10.1109/DAC18074.2021.9586121
ArXiv
Slide
Top Conf. AI/ML Arch.
Qimera: Data-free Quantization with Synthetic Boundary Supporting Samples
Qimera: Data-free Quantization with Synthetic Boundary Supporting Samples
Kanghyun Choi, Deokki Hong, Noseong Park, Youngsok Kim, Jinho Lee
Neural Information Processing Systems (NeurIPS)   ·   Nov 2021   ·   arxiv:2111.02625
ArXiv
Slide
GitHub
Top Conf. AI/ML
Enabling Hard Constraints in Differentiable Neural Network and Accelerator Co-Exploration
Enabling Hard Constraints in Differentiable Neural Network and Accelerator Co-Exploration
Deokki Hong, Kanghyun Choi, Hye Yoon Lee, Joonsang Yu, Noseong Park, Youngsok Kim, Jinho Lee
Design Automation Conference (DAC)   ·   Jan 2023   ·   arxiv:2301.09312
ArXiv
Top Conf. AI/ML Arch.
It's All In the Teacher: Zero-Shot Quantization Brought Closer to the Teacher
It's All In the Teacher: Zero-Shot Quantization Brought Closer to the Teacher
Kanghyun Choi, Hye Yoon Lee, Deokki Hong, Joonsang Yu, Noseong Park, Youngsok Kim, Jinho Lee
Computer Vision and Pattern Recognition Conference (CVPR)   ·   Apr 2022   ·   arxiv:2203.17008
ArXiv
Slide
GitHub
Top Conf. AI/ML
Fast Adversarial Training with Dynamic Batch-level Attack Control
Fast Adversarial Training with Dynamic Batch-level Attack Control
Jaewon Jung, Jaeyong Song, Hongsun Jang, Hyeyoon Lee, Kanghyun Choi, Noseong Park, Jinho Lee
Design Automation Conference (DAC)   ·   Jul 2023   ·   10.1109/DAC56929.2023.10247930
Top Conf. AI/ML

Teaching Experience