Meitu launched the "self-developed AI visual model" version 3.0, which claims to be "fully applied to imaging and design products".

Thanks to the CTOnews.com netizen Mr. Air, Xiao Zhanche's clue delivery! CTOnews.com October 10 news, according to Meitu official account, Meitu held a "15th anniversary birthday party" yesterday, Meitu CEO Wu Xinhong released the self-developed AI visual model MiracleVision (fantasy intelligence) 3.0 version.

▲ source Meitu Meitu claimed that 100 days after its launch, Meitu AI visual model MiracleVision3.0 will be fully applied to Meitu's image and design products, and will help e-commerce, advertising, games, animation, film and television five major industries.

Wu Xinhong said that three months ago, the level of MiracleVision painting was still in its infancy, but now version 3.0 can depict real and delicate picture details, and users can now visit the official website of AI visual creation tool "WHEE".

▲ image source Meitu CTOnews.com found that Meitu disassembled the core competence of the self-developed AI visual model MiracleVision into two major features: "fantastic ideas" and "intelligent creation":

At the "fantastic idea" level, MiracleVision claims to lower the threshold of public use through the function of "intelligent association of prompt words". When users enter keywords, MiracleVision can automatically supplement relevant expressions, such as light and shadow effect, texture, style, picture quality and so on.

This level has the function of "precise control of prompt words", which can meet the "more professional design requirements", such as using descriptions such as "close range", "vision", "smooth light" and "backlight" to control the final generation effect.

At the level of "intelligent creation", MiracleVision claims that it can further enrich the details and enhance the expression of the work through the function of "deepening creation". Through the "AI picture expansion" function to make the work larger size and richer details; but also through the "local modification" function, part of the picture can be accurately modified and adjusted, and "resolution enhancement" function to generate high-definition images.

Wu Xinhong said that MiracleVision will give full play to its ability of ingenious ideas and intelligent creation to work deeply in the above five industries and continue to help more design scenes, and it is expected that the popularity of relevant visual models will reach maturity in 2026-2030.

Meitu, source of ▲ image