Shulou(Shulou.com)11/24 Report--

CTOnews.com November 21 news, the underlying vision is a class of scenes in the field of computer vision, its algorithm can perform visual signal processing at the pixel level, complete the restoration of images and videos, quality improvement, style transformation tasks.

Huawei Noah's Ark Lab has created MindSpore Editing low-level vision suite based on Shengsi MindSpore AI framework, implementing and uniformly designing SOTA algorithm and model for mainstream low-level vision tasks, providing easy-to-use training and reasoning interfaces.

MindSpore Editing has officially released version 1.0 in the Shengsi open source community and is open to developers.

According to introduction, MindSpore Editing bottom vision suite adheres to the concept of simplicity and easy-to-use, adopts layered architecture design, encapsulates API for developers to call with one button, supports typical bottom vision tasks such as super-division, de-noise, de-blurring, repair, video frame insertion, rain removal, etc., and end-to-end supports data processing, training, fine-tuning, deployment, reasoning and evaluation functions of each type of task.

▲ Video Superscore

▲ Video frame insertion CTOnews.com with interface/code address:

https://github.com/mindspore-lab/mindediting

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.