On May 20, 2019, the Nihon Keizai Shimbun newspaper and the Nikkei e-edition published an article on "Skelton-Net," a technology in which Couger's AI team participated and whose paper was accepted for publication.
Click here for the article in the Nikkei e-edition.
New Methodology for Advanced Recognition Technology Expected to be Widely Applied to Automated Driving Cars, Robots, and Other Applications
The "Pixel SkelNetOn" competition in which the two members from Couger participate,d competes for the accuracy of AI models that recognize skeletons from shape images, and generally uses a data-generating algorithm called a GAN to extract the skeleton.
In contrast, our members from Couger developed a proprietary algorithm called "Skeleton-Net," which combines a method for understanding images at the pixel level called "U-Net" and a method for extracting the shape of an image with high accuracy called "HED," and succeeded in significantly improving the accuracy of skeletal extraction. (improved from 62% to 77%).
In the future, Couger will promote the application of its Human-like AI assistant "Virtual Human Agent," which it is still in development, to non-verbal communication.
