正在加载视频...

视频加载失败

I'm excited to share our new work Diffusion as Shader (DaS), a versatile controllable video generation method for various tasks: object manipulation, camera control, mesh-to-video, and motion transfer. Project page: Github:

33,363 次观看 • 1 年前 •via X (Twitter)

4 条评论

Yuan Liu 的头像
Yuan Liu1 年前

Paper: The key idea is to adopt the 3D tracking points as control signals in video generation. By manipulating the 3D tracking points, we achieve various types of video generation control.

Yuan Liu 的头像
Yuan Liu1 年前

We believe that future video diffusion models could have more 3D awareness with 3D representations inside, which would not only improve consistency but also make it more controllable. Our model, DaS, takes a step in this direction.

UserInterface 的头像
UserInterface4 年前

Need Professional Video Production, Music Videos, Commercials, Graphic Design, or Photo Retouching? We will take your project from concept to completion. #services #creative #DMV

Bilawal Sidhu 的头像
Bilawal Sidhu1 年前

Very cool work, congrats!

相关视频