This inventor holds 1 USPTO granted patent. Top assignee: Yonsei University. Active years: 2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2026
Title: Jae Hyeon Moon and Neural Network Quantization
Introduction
Jae Hyeon Moon is an inventor based in Seoul, South Korea, and is associated with Yonsei University. His patent work concerns efficient quantization for artificial neural networks and image processing.
Latest Patents
Moon is named on 1 patent: “Quantization Apparatus and Method for Artificial Neural Network and Image Processing Device Having the Same.” The invention describes dividing a neural-network input matrix into channels, analyzing the distribution of values in each channel, and selecting a suitable quantizer for each one. This approach aims to improve computational efficiency while limiting performance loss caused by quantization errors.
Career Highlights
Moon’s listed patent reflects work at the intersection of neural-network computation and image processing. The disclosed method adapts quantization to the value distributions within individual channels rather than applying a single quantizer uniformly.
Collaborations
The patent lists Bum Sub Ham and Do Hyung Kim as collaborators. Moon is associated with Yonsei University.
Conclusion
Jae Hyeon Moon’s patent focuses on channel-specific quantization as a way to make neural-network and image-processing operations more efficient while preserving performance.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile