StableV2V: Stablizing Shape Consistency in Video-to-Video Editing
Chang Liu, Rui Li, Kaidong Zhang, Yunwei Lan, Dong Liu
[Paper
] / [Project
] / [GitHub
] / [DAVIS-Edit (HuggingFace)
] / [Models (wisemodel)
] / [DAVIS-Edit (wisemodel)
] / [Models (ModelScope)
] / [DAVIS-Edit (ModelScope)
]
Official pre-trained model weights of the paper titled "StableV2V: Stablizing Shape Consistency in Video-to-Video Editing".
Model Weights Structure
We construct our model weights following the structure below:
StableV2V
βββ controlnet-depth <----- ControlNet (depth), required by CIG
βββ controlnet-scribble <----- ControlNet (scribble, needed in sketch-based editing application)
βββ ctrl-adapter-i2vgenxl-depth <----- Ctrl-Adapter (I2VGen-XL, depth), required by CIG
βββ i2vgenxl <----- I2VGen-XL, required by CIG
βββ instruct-pix2pix <----- InstructPix2Pix, required by PFE
βββ paint-by-example <----- Paint-by-Example, required by PFE
βββ stable-diffusion-v1-5-inpaint <----- SD Inpaint, required by PFE
βββ stable-diffusion-v1.5 <----- SD v1.5, required by CIG
βββ 50000.ckpt <----- Shape-guided depth refinement network
βββ README.md
βββ dpt_swin2_large_384.pt <----- MiDaS, required by ISA
βββ raft-things.pth <----- RAFT, required by ISA
βββ u2net.pth <----- U2-net, required by ISA
βββ 50000.ckpt <----- Shape-guided depth refinement network, required by ISA
- Downloads last month
- 82