AI Video Model

Vidu is China's first long-duration, highly consistent, and highly dynamic video model released by Shengshu Technology and Tsinghua University. The model uses the original Diffusion and Transformer fusion architecture U-ViT, which supports one-click generation of up to 16 seconds of high-definition video content with a resolution of up to 1080P. Vidu can not only simulate the real physical world, but also has rich imagination, multi-lens generation, and high spatiotemporal consistency. Vidu function Long-duration high-definition video generation: Vidu can generate up to 16 seconds of high-definition video based on the user's text description...

Recently, Picsart AI Research and other teams jointly released an AI video model called StreamingT2V, which can generate videos up to 1200 frames and 2 minutes long, which technically surpasses the previously popular Sora model. The release of StreamingT2V not only made a breakthrough in video length, but it is also a free and open source project that can be seamlessly compatible with models such as SVD and animatediff, which is of great significance to the development of the open source ecosystem. Before Sora, the video…

Meta’s GenAI team introduced a video-to-video synthesis model called Fairy that is faster and more temporally consistent than existing models. The research team demonstrated Fairy’s performance in several applications, including character/object replacement, stylization, and long-form video generation. For example, simple text prompts such as “in the style of Van Gogh” are enough to edit the source video. For example, the text command “turn into a snowman” turns the astronaut in the video into a snowman. Fairy’s visual coherence is a particularly challenging problem because the same prompts can be used to generate different videos.