Scaling

class plugins.convert.scaling.sharpen.Scaling(config_file=None)

Bases: Adjustment

Sharpening Adjustments for the face applied after warp to final frame

Methods Summary

`box`(new_face, kernel_size, radius, amount)	Sharpen using box filter
`gaussian`(new_face, kernel_size, radius, amount)	Sharpen using gaussian filter
`get_kernel_size`(new_face, radius_percent)	Return the kernel size and central point for the given radius
`process`(new_face)	Sharpen using the requested technique
`run`(new_face)	Perform selected adjustment on face
`unsharp_mask`(new_face, kernel_size, center, ...)	Sharpen using unsharp mask

Methods Documentation

classmethod box(new_face: ndarray, kernel_size: tuple[int, int], radius: int, amount: float) → ndarray

Sharpen using box filter

Parameters:

new_face (numpy.ndarray) – The batch of swapped image patches that is to have sharpening applied
kernel_size (tuple[int, int]) – The sharpening kernel size
radius (int) – The pixel radius the kernel
amount (float) – The amount of sharpening to apply

Returns:

The batch of swapped faces with box sharpening applied

Return type:

numpy.ndarray

classmethod gaussian(new_face: ndarray, kernel_size: tuple[int, int], radius: float, amount: float) → ndarray

Sharpen using gaussian filter

Parameters:

new_face (numpy.ndarray) – The batch of swapped image patches that is to have sharpening applied
kernel_size (tuple[int, int]) – The sharpening kernel size
radius (int) – The pixel radius the kernel. Unused
amount (float) – The amount of sharpening to apply

Returns:

The batch of swapped faces with gaussian sharpening applied

Return type:

numpy.ndarray

classmethod get_kernel_size(new_face: ndarray, radius_percent: float) → tuple[tuple[int, int], int]

Return the kernel size and central point for the given radius: relative to frame width.

Parameters:

new_face (numpy.ndarray) – The swapped image patch that is to have sharpening applied
radius_percent (float) – The percentage of the image size to use as the sharpening kernel

Returns:

kernel_size (tuple[int, int]) – The sharpening kernel
radius (int) – The pixel radius the kernel

Return type:

tuple[tuple[int, int], int]

process(new_face: ndarray) → ndarray

Sharpen using the requested technique

Parameters:: new_face (numpy.ndarray) – A batch of swapped image patch that is to have sharpening applied
Returns:: The batch of swapped faces with sharpening applied
Return type:: numpy.ndarray

run(new_face): Perform selected adjustment on face

classmethod unsharp_mask(new_face: ndarray, kernel_size: tuple[int, int], center: float, amount: float) → ndarray

Sharpen using unsharp mask

Parameters:

new_face (numpy.ndarray) – The batch of swapped image patches that is to have sharpening applied
kernel_size (tuple[int, int]) – The sharpening kernel size
radius (int) – The pixel radius the kernel. Unused
amount (float) – The amount of sharpening to apply
center (float)

Returns:

The batch of swapped faces with unsharp-mask sharpening applied

Return type:

numpy.ndarray

classmethod box(new_face: ndarray, kernel_size: tuple[int, int], radius: int, amount: float) → ndarray

Sharpen using box filter

Parameters:

new_face (numpy.ndarray) – The batch of swapped image patches that is to have sharpening applied
kernel_size (tuple[int, int]) – The sharpening kernel size
radius (int) – The pixel radius the kernel
amount (float) – The amount of sharpening to apply

Returns:

The batch of swapped faces with box sharpening applied

Return type:

numpy.ndarray

classmethod gaussian(new_face: ndarray, kernel_size: tuple[int, int], radius: float, amount: float) → ndarray

Sharpen using gaussian filter

Parameters:

new_face (numpy.ndarray) – The batch of swapped image patches that is to have sharpening applied
kernel_size (tuple[int, int]) – The sharpening kernel size
radius (int) – The pixel radius the kernel. Unused
amount (float) – The amount of sharpening to apply

Returns:

The batch of swapped faces with gaussian sharpening applied

Return type:

numpy.ndarray

classmethod get_kernel_size(new_face: ndarray, radius_percent: float) → tuple[tuple[int, int], int]

Return the kernel size and central point for the given radius: relative to frame width.

Parameters:

new_face (numpy.ndarray) – The swapped image patch that is to have sharpening applied
radius_percent (float) – The percentage of the image size to use as the sharpening kernel

Returns:

kernel_size (tuple[int, int]) – The sharpening kernel
radius (int) – The pixel radius the kernel

Return type:

tuple[tuple[int, int], int]

process(new_face: ndarray) → ndarray

Sharpen using the requested technique

Parameters:: new_face (numpy.ndarray) – A batch of swapped image patch that is to have sharpening applied
Returns:: The batch of swapped faces with sharpening applied
Return type:: numpy.ndarray

classmethod unsharp_mask(new_face: ndarray, kernel_size: tuple[int, int], center: float, amount: float) → ndarray

Sharpen using unsharp mask

Parameters:

new_face (numpy.ndarray) – The batch of swapped image patches that is to have sharpening applied
kernel_size (tuple[int, int]) – The sharpening kernel size
radius (int) – The pixel radius the kernel. Unused
amount (float) – The amount of sharpening to apply
center (float)

Returns:

The batch of swapped faces with unsharp-mask sharpening applied

Return type:

numpy.ndarray