Beautiful picturesThe company announced today that its "Fantasy Large Model"The video generation capability has been fully upgraded, and on the basis of the triple advancement in generation capability, generation efficiency and model performance, combined with a number of Meitu's self-developed technologies in the field of computer vision, video generation has been realized.Duration, picture quality, smoothness, authenticity and believabilityand other enhancements.
It is reported that the length of a single text-generated video and a single map-generated video of the Meitu Wonder Big Model is up to 5 seconds, and it supports the generation of 1 minute, 24FPS, 1080P The upgraded video generation function will gradually cover Meitu Xiu Xiu, Beauty Camera, Wink, Open Shoot, Meitu Design Studio, WHEE, MOKI and other products.
On January 2 this year, MiracleVision, the self-developed AI visual model of Mito, passed the Interim Measures for the Administration of Generative Artificial Intelligence Services and will be open to the public. MiracleVision was tested in June 2023, providing AI model capabilities for Meitu Xiu Xiu, Meiyan Camera, Wink, Meitu Design Studio, WHEE, Meitu Cloud Repair and other products, and helping Meitu build an AI product ecosystem constructed by the bottom layer, middle layer and application layer.