Systems and methods for selective enhancement of skin features in images
AssigneeShinyfields Limited
InventorsDmitry Sytnik, Andrey Frolov
Life of this patent
Estimating the term…
Abstract
Exemplary embodiments are directed to a system for selective enhancement of skin features in an image. The system includes an interface configured to receive as input an original image, and a processing device in communication with the interface. The processing device is configured to process the original image using a neural network to detect one or more skin imperfections in the original image, generate a neural network mask of the original image for the one or more skin imperfections in the original image, generate one or more source patches based on the original image, and, replace the one or more skin imperfections in the original image with the one or more source patches to generate a patched skin image.
Claims20
A method for selective enhancement of skin features in an image, the method comprising:
- receiving as input at an interface an original image;
- processing the original image using a neural network to detect one or more skin imperfections in the original image;
- generating a neural network mask of the original image for the one or more skin imperfections in the original image;
- generating one or more source patches based on the original image;
- replacing the one or more skin imperfections in the original image with the one or more source patches to generate a patched skin image;
- generating a masked skin image including a skin mask;
- and
- generating a blurred image, the blurred image including blurring of the skin encompassed by the skin mask without affecting facial feature details.
A non-transitory computer-readable medium storing instructions at least for selective enhancement of skin features in an image that are executable by a processing device, wherein execution of the instructions by the processing device causes the processing device to:
- receive as input at an interface an original image;
- process the original image using a neural network to detect one or more skin imperfections in the original image;
- generate a neural network mask of the original image for the one or more skin imperfections in the original image;
- generate one or more source patches based on the original image;
- replace the one or more skin imperfections in the original image with the one or more source patches to generate a patched skin image;
- generate a masked skin image including a skin mask;
- and
- generate a blurred image, the blurred image including blurring of the skin encompassed by the skin mask without affecting facial feature details.
Dependencies are inferred from the claim text.