lqhl/realesrgan 🖼️🔢❓✓ → 🖼️

▶️ 139.7K runs 📅 Oct 2023 ⚙️ Cog 0.8.6 🔗 GitHub 📄 Paper ⚖️ License

image-restoration image-upscaling

About

Image restoration and face enhancement

Example Output

Output

Performance Metrics

12.47s Prediction Time

70.43s Total Time

All Input Parameters

{
  "img": "https://replicate.delivery/pbxt/JeM4zZQu0BGlvPiOqu6BclFQLcr4uCsRBulTS5mMVDqZiAdC/e512564400014599b143497df8fca4aa.jpeg",
  "tile": 0,
  "scale": 2,
  "version": "General - RealESRGANplus",
  "face_enhance": true
}

Input Parameters

img (required) Type: string: Input
tile Type: integerDefault: 0: Tile size. Default is 0, that is no tile. When encountering the out-of-GPU-memory issue, please specify it, e.g., 400 or 200
scale Type: numberDefault: 2: Rescaling factor
version Default: General - v3: RealESRGAN version. Please see [Readme] below for more descriptions
face_enhance Type: booleanDefault: false: Enhance faces with GFPGAN. Note that it does not work for anime images/vidoes

Output Schema

Output

Type: string • Format: uri

Example Execution Logs

img: /tmp/tmpvb852ci6e512564400014599b143497df8fca4aa.jpeg. version: General - RealESRGANplus. scale: 2.0. face_enhance: True. tile: 0.
Downloading: "https://github.com/xinntao/facexlib/releases/download/v0.1.0/detection_Resnet50_Final.pth" to /src/gfpgan/weights/detection_Resnet50_Final.pth
  0%|          | 0.00/104M [00:00<?, ?B/s]
 26%|██▋       | 27.5M/104M [00:00<00:00, 288MB/s]
 53%|█████▎    | 55.0M/104M [00:00<00:00, 228MB/s]
 74%|███████▍  | 77.5M/104M [00:00<00:00, 221MB/s]
 95%|█████████▍| 98.9M/104M [00:00<00:00, 215MB/s]
100%|██████████| 104M/104M [00:00<00:00, 225MB/s]
Downloading: "https://github.com/xinntao/facexlib/releases/download/v0.2.2/parsing_parsenet.pth" to /src/gfpgan/weights/parsing_parsenet.pth
  0%|          | 0.00/81.4M [00:00<?, ?B/s]
 16%|█▌        | 12.6M/81.4M [00:00<00:00, 132MB/s]
 49%|████▉     | 39.7M/81.4M [00:00<00:00, 221MB/s]
 80%|███████▉  | 64.8M/81.4M [00:00<00:00, 241MB/s]
100%|██████████| 81.4M/81.4M [00:00<00:00, 249MB/s]
/root/.pyenv/versions/3.8.18/lib/python3.8/site-packages/torch/nn/functional.py:3103: UserWarning: The default behavior for interpolate/upsample with float scale_factor changed in 1.6.0 to align with other frameworks/libraries, and now uses scale_factor directly, instead of relying on the computed output size. If you wish to restore the old behavior, please set recompute_scale_factor=True. See the documentation of nn.Upsample for details.
warnings.warn("The default behavior for interpolate/upsample with float scale_factor changed "

Version Details

Version ID: 56f2be05920413e1189c32a5fb2f767b357187c887d67114cace11c18d86ab49
Version Created: October 5, 2023

Run on Replicate →