近日，Meta推出了一款全新的AI视频生成模型FlowVid。该模型由得克萨斯大学奥斯汀分校和Meta GenAI团队共同研发，提出了一种能够保持一致性的V2V（视频到视频）合成框架。据悉，FlowVid仅需1.5分钟就能生成一段4秒，每秒30帧、分辨率为512×512的视频。
FlowVid的出现，不仅为视频生成提供了新的可能性，还无缝地与现有的I2I模型配合，支持多种修改方式，包括风格化、物体替换和局部编辑等。这无疑为视频创作提供了更为广阔的空间，也为用户带来了更为丰富的视觉体验。
据相关人士透露，FlowVid的推出，标志着Meta在AI视频生成领域的一次重要突破。未来，我们有理由相信，Meta还会在这个领域带来更多的创新和惊喜。
英文标题：Meta Unveils AI Video Generation Model FlowVid
关键词：AI video generation, FlowVid, Meta
News content:
Recently, Meta has launched a brand new AI video generation model called FlowVid. Developed by the University of Texas at Austin and the Meta GenAI team, this model proposes a consistent V2V (video to video) synthesis framework. It is capable of generating a 4-second video with 30 frames per second and a resolution of 512×512 within just 1.5 minutes.
FlowVid not only opens up new possibilities for video generation but also seamlessly integrates with existing I2I models, supporting various editing methods such as stylization, object replacement, and local editing. This expands the scope of video creation and enhances the visual experience for users.
According to insiders, the introduction of FlowVid represents an important breakthrough for Meta in the field of AI video generation. We have reason to believe that Meta will bring more innovation and surprises in this area in the future.
【来源】https://mp.weixin.qq.com/s/06Tpe6a4Q3wkPyNX01PyFg
Views: 1