yuval-alaluf/restyle_encoder 🖼️❓🔢✓ → 🖼️

▶️ 99.7K runs 📅 Sep 2021 ⚙️ Cog 0.4.1 🔗 GitHub 📄 Paper ⚖️ License

image-style-transfer image-to-image

About

ReStyle: A Residual-Based StyleGAN Encoder via Iterative Refinement

Example Output

Output

[object Object]

Performance Metrics

74.05s Total Time

All Input Parameters

{
  "input": "https://replicate.delivery/mgxm/24ebe659-33d0-4fec-a4b2-3b99f7960744/zuckerberg.jpg",
  "encoding_type": "toonify",
  "num_iterations": "2",
  "display_intermediate_results": false
}

Input Parameters

input (required) Type: string: Path to input image
encoding_type (required): Which domain you wish to run on
num_iterations Type: integerDefault: 5Range: 1 - 10: Number of ReStyle iterations to run. For `faces` we recommend 5 iterations and for `toonify` we recommend 1 to 2 iterations.
display_intermediate_results (required) Type: boolean: Whether to display all intermediate outputs. If unchecked, will display only the final result.

Output Schema

Output

Type: string • Format: uri

Example Execution Logs

Loading faces model...
Loading ReStyle pSp from checkpoint: pretrained_models/restyle_psp_ffhq_encode.pt
Done!
Loading toonify model...
Loading ReStyle pSp from checkpoint: pretrained_models/restyle_psp_toonify.pt
Done!
Aligning image...
Done!
Running inference...
Finished inference in 0.47192955017089844 seconds!
Preparing result...

Version Details

Version ID: c677ac3e9f0dd4a8c8947094865cfa5c2cf252eab5822cbaf649ca08671f9afd
Version Created: September 29, 2022

Run on Replicate →