Kazuki Kanamoto, a member of the Imaging Group in Shimadzu’s Analytical & Measuring Instruments Division R&D Department, participated individually in the Image Matching Challenge 2025 on Kaggle, the world’s largest AI competition platform. Competing against 943 teams, Kanamoto achieved 6th place finish, earning a gold medal. This marks his second gold medal, as he previously won one in 2024, and he has been recognized as a Kaggle Master, ranking in the top 1% of participants.
Kazuki Kanamoto, Imaging Group, R&D Department, Analytical & Measuring Instruments Division
On Kaggle, over 20 million data scientists and AI engineers compete to develop the most accurate AI models for challenges posed by companies and researchers. The Image Matching Challenge 2025 was held from April 2 to June 3 as part of a workshop at the Conference on Computer Vision and Pattern Recognition (CVPR), the world’s leading conference in the field of computer vision. The challenge was to “accurately reconstruct a 3D scene from a set of images for each target.” Kanamoto used two models, “detecting feature points” and “matching key points between images,” and devised image preprocessing and procedures to ensure that the models functioned appropriately in order to achieve high-precision 3D scene reconstruction.
As an image processing engineer, Kanamoto has contributed to the development of several Shimadzu products, including the PLUS ALGO AI-powered multi-data analysis platform, the PLUS ALGO AM X-ray fluoroscopy software, and the Cell Pocket cell observation support application. Shimadzu Corporation remains committed to contributing to customers’ R&D and productivity by leveraging AI for integrated analysis of instrument data, image recognition, data visualization, and interpretive support.
Related links
Kaggle
Image Matching Challenge 2025
January 30, 2025 | News
Shimadzu Employee wins Gold Medal in Kaggle, the World’s Largest AI Competition
Promoted to “Kaggle Master”, placing him in the top 1% of Contestants
December 13, 2021 | Press Release
Shimadzu Releases New Version of Its Cell Pocket Web Application for Supporting Cell Observation
Cell Image Management and Analysis Software Augmented by AI (Deep Learning)
