Meitu releases 7 AIGC products, including visually appealing large-scale models with a focus on aesthetics

Meitu releases 7 AI graphics products with a focus on aesthetics

【PhoneAuto News】On June 19th, Meitu held the second Image Festival with the theme of “AI Era Image Productivity Tools” and released seven new products: AI visual creation tool WHEE; AI voice-over video tool ; desktop AI video editing tool WinkStudio; Meitu Design Studio 2.0, which focuses on AI commercial design; AI digital human generation tool DreamAvatar; Meitu AI assistant RoboNeo; and Meitu visual model MiracleVision.

Meitu founder, chairman and CEO Wu Xinhong said, “In the past, Meitu mainly served life scenes, such as shooting, retouching, video editing, and social sharing. In the AI era, Meitu will provide more productivity scene services in addition to life scenes.”

• Sony Renames IMX and Launches New Mobile Sensor Brand LYTIA
• Huawei charges patent fees from 30 Japanese companies Ren Zhengfei once said that he would collect patent fees when he has free time
• Sales of foldable screen smartphones during the 618 shopping festival have increased more than three times, and prices have been refreshed to a new low!

At the Image Festival, Meitu released the first domestic AI visual model MiracleVision that understands aesthetics. Wu Xinhong explained that Meitu integrates the long-term accumulated aesthetic cognition into the MiracleVision visual model and builds an aesthetic evaluation system based on machine learning, which gives the model-generated results an “aesthetic score” and continuously improves the model’s understanding of aesthetics. In addition, the MiracleVision visual model has a unique aesthetic creator ecology, with professional personnel with deep aesthetic backgrounds such as artists and designers participating in the construction of the visual model together.

Unlike other visual models on the market, MiracleVision visual model has strong visual expression and creativity, and reverse-pushes technology evolution from visual creation scenes such as painting, design, film and television, photography, games, 3D, animation, and is good at generating directions such as Asian portrait photography, Chinese style, commercial design, etc.

Wu Xinhong stated that a good visual large model needs to be supported by a good “model ecosystem”, and this process cannot be separated from the continuous participation of creators and developers. Meitu will systematically build the model ecosystem through model training, distribution, and commercialization.

We will continue to update Phone&Auto; if you have any questions or suggestions, please contact us!