W.A.L.T
Stanford UniversityGoogle ResearchGeorgia Institute of TechnologyVideo generationText-to-video
W.A.L.T is video generation model published by Stanford University,Google Research,Georgia Institute of Technology in 2023 featuring 4719000000.0 parameters.
About W.A.L.T
We present W.A.L.T, a transformer-based approach for photorealistic video generation via diffusion modeling. Our approach has two key design decisions. First, we use a causal encoder to jointly compress images and videos within a unified latent space
Details
- Provider
- Stanford University,Google Research,Georgia Institute of Technology
- Task
- Video generation,Text-to-video
- Parameters
- 4719000000.0
- Released
- 2023-12-11
- Open weights
- No