A Face Data Identity De-recognition Method Based on Generative Adversarial Networks

Abstract

The invention discloses a face data identity de-recognition method based on a generative confrontation network, and relates to the technical field of biometric feature recognition and artificial intelligence security. The attribute feature encoding and expression pose encoding of the face image are combined to obtain the first latent vector, the second latent vector is obtained through the mapping network, and the second latent vector is sent to the generation network to obtain the output image, and the face image attribute feature and expression are completed. The fusion of poses, using face-changing technology, realizes the identity de-identification of human vision; secondly, constructs an adversarial vector mapping network, takes the second hidden vector and inputs it into the adversarial vector mapping network to obtain the adversarial hidden vector, and then obtains face recognition through the generation network For adversarial sample images with large gaps in model recognition results and small gaps in human vision, the identity de-identification of the face recognition model can be realized through adversarial sample technology.

Description

technical field [0001] The invention relates to the technical fields of biological feature recognition and artificial intelligence security, in particular to a face data identity de-recognition method based on a generative confrontation network. Background technique [0002] In the era of artificial intelligence based on big data training, computer vision technology is widely used in security and tracking task scenarios, but the resulting security problems of misuse of face data have aroused people's concern about the necessity of protecting face privacy. Pay attention to. Traditional anonymized face technologies, such as mosaicing or blurring faces, have the disadvantage of large information loss, making it impossible for users or data developers to effectively use the anonymized face data. With the introduction of the concepts of adversarial samples and generative adversarial networks, two face recognition (De-identify, De-id) technologies represented by adding random noi...

AI Technical Summary

Problems solved by technology

However, the disturbance of random noise will cause the image quality to degrade, and the face-changing technology will be at the cost of completely losing the original face information.

Images