At the China AIGC Industry Summit held on April 17, Wu Xinhong, founder of Meitu, said that there are three key points of competition for major domestic video models, namely creativity beyond reality, workflow integration, and vertical scenarios.
At the beginning of this year, Sora's debut took the big video model a new step, and also stimulated technology companies to step up technical research and increase resource investment in the big video model. It is expected that quite a few “domestic Sora” may be launched in the second half of the year.
According to Wu Xinhong, first, the big video model needs to generate images that transcend reality, complement real shots, and become a new way to achieve special effects. Second, video generation has limited application scenarios and needs to be combined with existing AI capabilities and traditional video technology to form a workflow. Third, the “usability” of video in vertical scenes such as e-commerce, advertising, gaming, animation, film and television is critical.
“The big video model represented by Sora is just the beginning.” Wu Xinhong said that future video models can be more closely integrated with video production workflows, and subsequent 1-5 minute videos can be generated, which is worth the expectations of users and the industry.
At the event, Wu Xinhong showed the guests a 60-second video. The video was completed in collaboration with products such as Wink, WHEE, and Kickoff from the Meitu product family bucket. By optimizing processes such as story scripting, video generation, and video editing, Wu Xinhong showed the video generation capabilities of Meitu's fantastic model Miracle Vision.
Wu Xinhong revealed that the Meitu Video Big Model is undergoing a technical upgrade from 1.0 to 2.0. At the Meitu Video Festival in June of this year, new big video models and new productivity tools will be unveiled.