rinongal/stylegan-nada 🖼️📝❓✓ → 🖼️

▶️ 102.5K runs 📅 Oct 2021 ⚙️ Cog 0.4.1 🔗 GitHub 📄 Paper ⚖️ License

image-style-transfer image-to-image image-to-video

About

StyleGAN-NADA: CLIP-Guided Domain Adaptation of Image Generators

Example Output

Output

[object Object]

Performance Metrics

8.00s Total Time

All Input Parameters

{
  "input": "https://replicate.delivery/mgxm/e348c7fa-341d-443a-b56c-67ba6a180d0b/emma.jpg",
  "style_list": "joker,anime,modigliani",
  "output_style": "all",
  "video_format": "mp4",
  "with_editing": true,
  "generate_video": false
}

Input Parameters

input (required) Type: string: input image
style_list Type: stringDefault: joker,anime,modigliani: Comma separated list of models to use. Only accepts models from the output_style list. Will only be used if the chosen output_style is list
output_style Default: all: Which output style do you want to use? Select 'all' to generate a collage.
video_format Default: mp4: Choose gif to display in browser, mp4 for a higher-quality downloadable video
with_editing Type: booleanDefault: true: Apply latent space editing to the generated video
generate_video Type: booleanDefault: false: Generate a video instead of an output image. If more than one style is used, will interpolate between styles.

Output Schema

Output

Type: string • Format: uri

Example Execution Logs

Aligned image has shape: (256, 256)

Version Details

Version ID: 6b2af4ac56fa2384f8f86fc7620943d5fc7689dcbb6183733743a215296d0e30
Version Created: September 29, 2022

Run on Replicate →