Docs | License | Installation | Model Zoo
We provide a wide range of pretrained imaginaire models for different tasks. All the models were trained using an NVIDIA DGX 1 machine with 8 32GB V100 using NVIDIA PyTorch docker 20.06 or later.
Algorithm Name | Task | Model | Resolution |
---|---|---|---|
pix2pixHD | Cityscapes, segmentation to image | download | 1024x512 |
SPADE | COCO-Stuff, segmentation to image | download | 256x256 |
Algorithm Name | Task | Model | Resolution |
---|---|---|---|
UNIT | Winter <-> Summer | download | 256x256 |
MUNIT | AFHQ Dog <-> Cat | download | 256x256 |
FUNIT | AniamlFaces | download | 256x256 |
COCO-FUNIT | AniamlFaces | download | 256x256 |
COCO-FUNIT | Mammals, Full body animal translation | download | 256x256 |
Algorithm Name | Task | Model | Resolution |
---|---|---|---|
vid2vid | Cityscapes, segmentation to video | download | 1024x512 |
fs-vid2vid | FaceForensics, landmarks to video | download | 512x512 |
Algorithm Name | Task | Model | Resolution |
---|---|---|---|
wc-vid2vid | Cityscapes, segmentation to video | download | 1024x512 |
GANcraft | Semantic block worlds to realistic worlds | download | arbitrary |