RaonPeople 'NAVI AI 4.0 Release... "Defect Detection with Deep Learning Vision Inspection"'
[Asia Economy Reporter Hyunseok Yoo] AI specialist company Raonpeople announced on the 22nd the release of a new version of ‘NAVI AI 4.0,’ which detects defects through good product deep learning vision inspection.
In NAVI AI 4.0, the functions of the existing three inspection modules?MARS, MERCURY, and VENUS?have been enhanced, and a new unsupervised learning module called JUPITER was added. Additionally, a new service was launched that allows defect detection by learning only good products without the need to train on various unspecified defects.
NAVI AI 4.0 also enables automatic classification and label generation when additional data is inserted through the latest Active Learning techniques. It has been upgraded to allow ultra-high-speed inspection of 100 Mega Pixels per second per GPU.
With improved accuracy of multi-label learning that trains multiple types of defects in a single image at once, it is optimized for detecting various defects. The Patch Train function, which cuts specific areas to train large images (over 100M resolution) that are difficult to learn, has been enhanced for a higher level of sophistication. Besides supporting deep learning technology specialized for manufacturing, inspection speed has dramatically improved through support for the latest Cuda11, TensorRT, and OpenVINO, and efficient use of GPU and CPU resources is also an advantage, according to the company.
Moreover, the service supports multiple languages including English, Chinese, and Japanese, and provides convenience for customers and managers through 15 updated diverse augmentations.
Hot Picks Today
As Samsung Falters, Chinese DRAM Surges: CXMT Returns to Profit in Just One Year
- "Most Americans Didn't Want This"... Americans Lose 60 Trillion Won to Soaring Fuel Costs
- Man in His 30s Dies After Assaulting Father and Falling from Yongin Apartment
- Samsung Union Member Sparks Controversy With Telegram Post: "Let's Push KOSPI Down to 5,000"
- "Why Make Things Like This?" Foreign Media Highlights Bizarre Phenomenon Spreading in Korea
Choo Yeonhak, Director of Raonpeople, said, “We improved the configuration and user convenience so that it can be applied to manufacturing even if managers have no knowledge of deep learning, as well as providing an intuitive UI setting. By applying improvements requested by customers and various field experiences from previous versions, productivity will be further enhanced with accurate, fast, and cost-effective AI vision inspection.”
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
