Skip to content

index.md

Latest commit

 

History

History
216 lines (148 loc) · 5.08 KB

index.md

File metadata and controls

216 lines (148 loc) · 5.08 KB

Waifu2x

Waifu2x Index / Waifu2x

Auto-generated documentation for waifu2x module.

denoise_image

Show source in init.py:32

Remove noise from an image (src) using a scale model and an alpha model.

Arguments

  • args argparse.Namespace - argparse namespace containing config such as the block_size
  • src Image.Image - Pillow image to remove noise from
  • model Chain - model for scaling
  • should_print bool - Flag to enable print to console

Returns

  • Image.Image - Pillow image with noise removed

Signature

def denoise_image(
    args: argparse.Namespace, src: Image.Image, model: Chain, should_print: bool = True
) -> Image.Image: ...

load_models

Show source in init.py:157

Load models using a args config.

Arguments

  • args argparse.Namespace - argparse namespace containing config such as the arch and color

Returns

  • dict[str, Chain] - Mapping of model names to Chain models

Signature

def load_models(args: argparse.Namespace) -> dict[str, Chain]: ...

main

Show source in init.py:27

Run the program, from the main entry point.

Signature

def main() -> None: ...

run

Show source in init.py:211

Run waifu2x using mostly the same inputs as CLI ones.

Arguments

  • input_img_path str - Input image/ directory, defaults to "images/small.png"
  • output_img_path str - Directory to write output images to, defaults to "./"
  • gpu int - CUDA enabled GPU to use, defaults to -1 :param int | None quality: Set the quality of output images 1-100 (None=100), defaults to None :param str | None model_dir: Specify a custom directory containing models, defaults to None
  • scale_ratio float - Specify a scale, defaults to 2.0
  • batch_size int - description, defaults to 16
  • block_size int - description, defaults to 128
  • extension str - Select output extension png/webp
  • arch str - description, defaults to "VGG7"
  • method str - description, defaults to "scale"
  • noise_level int - description, defaults to 1
  • color str - description, defaults to "rgb"
  • tta_level int - description, defaults to 8
  • width int - description, defaults to 0
  • height int - description, defaults to 0
  • shorter_side int - description, defaults to 0
  • longer_side int - description, defaults to 0
  • should_print bool - description, defaults to True

Raises

  • ValueError - Output file extension not supported

Signature

def run(
    input_img_path: str = "images/small.png",
    output_img_path: str = "./",
    gpu: int = -1,
    quality: int | None = None,
    model_dir: str | None = None,
    scale_ratio: float = 2.0,
    batch_size: int = 16,
    block_size: int = 128,
    extension: str = "png",
    arch: str = "VGG7",
    method: str = "scale",
    noise_level: int = 1,
    color: str = "rgb",
    tta_level: int = 8,
    width: int = 0,
    height: int = 0,
    shorter_side: int = 0,
    longer_side: int = 0,
    should_print: bool = True,
) -> None: ...

split_alpha

Show source in init.py:128

Split the image into an rgb, and alpha tuple.

:param Image.Image src: image

Arguments

  • model Chain - model to use
  • should_print bool - print to stdout?, defaults to True

Returns

Type: tuple[Image.Image, Image.Image | None] rgb, and alpha tuple

Signature

def split_alpha(
    src: Image.Image, model: Chain, should_print: bool = True
) -> tuple[Image.Image, Image.Image | None]: ...

upscale_image

Show source in init.py:69

Upscale an image (src) using a scale model and an alpha model.

Arguments

  • args argparse.Namespace - argparse namespace containing config such as the scale_ratio and block size
  • src Image.Image - Pillow image to upscale
  • scale_model Chain - model to use for scaling
  • alpha_model Chain, optional - model to use for alpha. Defaults to None.
  • should_print bool - flag to enable output to console

Returns

  • Image.Image - upscaled Pillow image

Signature

def upscale_image(
    args: argparse.Namespace,
    src: Image.Image,
    scale_model: Chain,
    alpha_model: Chain | None = None,
    should_print: bool = True,
) -> Image.Image: ...

Modules