Kun He

Title: Kun He: Innovator in Feature Descriptor Matching

Introduction

Kun He is an accomplished inventor based in Kirkland, WA (US). He has made significant contributions to the field of image processing, particularly in feature descriptor matching. His innovative work has led to the development of a patented technology that enhances the accuracy of image recognition systems.

Latest Patents

Kun He holds a patent for a method of feature descriptor matching. This patent involves receiving a first input image and a second input image. A feature detector is utilized to identify features from both images. A descriptor extractor learns local feature descriptors based on a feature descriptor matching model trained with a ground truth data set. The system determines a listwise mean average precision (mAP) rank of candidate image patches from the second input image concerning a queried image patch from the first input image. The descriptor matcher generates a geometric transformation between the two images based on the listwise mAP and a convolutional neural network. This innovative approach enhances the efficiency and accuracy of image matching processes.

Career Highlights

Kun He is currently employed at Honda Motor Co., Ltd., where he applies his expertise in image processing and computer vision. His work at Honda focuses on integrating advanced technologies into automotive systems, improving safety and functionality.

Collaborations

Kun He collaborates with his coworker, Yan Lu, to further enhance their research and development efforts in the field of image processing. Their combined expertise contributes to innovative solutions within their projects.

Conclusion

Kun He is a notable inventor whose work in feature descriptor matching has the potential to revolutionize image recognition technologies. His contributions at Honda Motor Co., Ltd. and his patented innovations highlight his significant impact in the field.