On June 12, Meitu held the third "Meitu Image Festival" and released 6 products on site, covering the fields of commercial photography, oral broadcast videos, e-commerce design, design services, game marketing, and video generation. Wu Xinhong, founder of
Meitu, said that Meitu attempts to connect single-point AI functions into "AI workflow" to empower e-commerce, advertising, games, film and television, animation and other industries.
The so-called workflow is a general description of the workflow and the business rules between each operation step. Wu Xinhong believes that compared with traditional workflows, AI workflows have five major advantages: the integration of cloud + terminal computing power, one person is a team, easy selection of ideas, workflows can be planned arbitrarily, and self-iteration based on data models.
He predicted: "In a few years, generation will be more common than photography." Meitu's idea is: "First of all, we need to continuously improve the AI content of the product, and strongly associate the AI-driven functions with our positioning. To better drive the rapid growth of revenue and the continued explosion of users around the world.”
To this end, Meitu has updated some of its existing products, including the batch photo editing software “Meitu Cloud Repair” and AI broadcast video creation. software "Start Shooting", e-commerce design software "Meitu Design Studio"; in addition, it also released a new product design service platform "Zoku", a game advertising AI production and delivery platform "Qimi", and an AI video generation tool "moki".
(Source/Beautiful Picture)
It is worth mentioning that, unlike many Wensheng video models on the market that cover multiple types of generation needs at the same time, moki has chosen to focus on the AI short film creation track and will be launched on July 31. Wu Xinhong said that this was a decision made after comprehensive consideration of content controllability and production costs.
said that there are certain challenges in pursuing sora's technical capabilities and generation effects in the short term. Therefore, Meitu does not blindly compete head-on on the model side, but finds pain points and differentiated positioning on the application side.
In addition, Wu Xinhong also officially announced that Meitu’s self-developed AI visual model "miraclevision" has been upgraded to version v5.
In June 2023, Qixiang’s large model made its debut, becoming an important support for Meitu’s performance growth. According to official data, driven by AI, Meitu's global VIP membership has increased from 7.19 million in the same period last year to 10.63 million as of June 11, 2024.
(Source/Beautiful Picture)
(Author | Dong Wenshu, Editor | Li Yuan)
