User authentication using original and modified images
AssigneeInternational Business Machines Corporation
InventorsLuke Taher, Madeleine Frost, Vinh Tuan Thai, Paul Barry
Life of this patent
Estimating the term…
Abstract
Method and system are provided for user authentication using original and modified images. The method includes receiving an original image for a user, where the original image is a private image that meets certain configured criteria. The method uses a pre-trained Convolutional Neural Network (CNN) model to extract one or more image features of the original image and feeds the extracted image feature into a Generative Adversarial Network (GAN) image generator to realistically modify the extracted image feature to generate a modified image. The method authenticates a user based on recognition of a presented original image or a presented modified image.
Claims20
A computer-implemented method for user authentication using original and modified images, comprising:
- receiving an unaltered original image for a user, wherein the unaltered original image has not been previously shared;
- generating a modified image of the unaltered original image by applying a pre-trained Convolutional Neural Network (CNN) model to extract one or more image features of the unaltered original image, and feeding the one or more extracted image features into a Generative Adversarial Network (GAN) image generator to modify the one or more extracted image features to generate the modified image;
- and
- authenticating a user based on recognition of a presented unaltered original image or a presented modified image.
A system for user authentication using original and modified images, comprising:
- a processor and a memory configured to provide computer program instructions to the processor, the program instructions executable by the processor to cause the processor to:
- an image receiving component for receiving an original image for a user;
- receive an unaltered original image for a user, wherein the unaltered original image has not been previously shared;
- use a pre-trained Convolutional Neural Network (CNN) model to extract one or more image features of the unaltered original image;
- feed the one or more extracted image features into a Generative Adversarial Network (GAN) image generator to modify the one or more extracted image features to generate a modified image;
- and
- supply an image set of unaltered original and modified images to an authentication component for authenticating the user based on recognition of a presented unaltered original image or presented modified image.
A computer program product for user authentication using unaltered original and modified images, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
- receive an unaltered original image for a user, wherein the unaltered original image has not been previously shared;
- apply a pre-trained Convolutional Neural Network (CNN) model to extract one or more image features of the unaltered original image;
- feed the one or more extracted image features into a Generative Adversarial Network (GAN) image generator to modify the one or more extracted image features to generate a modified image;
- and
- authenticate the user based on recognition of a presented unaltered original image or a presented modified image.
Dependencies are inferred from the claim text.