Multi-modal identity authentication method based on vein similar image knowledge migration network

Abstract

The invention discloses a multi-modal identity authentication method based on a vein similar image knowledge migration network, a knowledge migration learning network model based on similar images anda supervision word bag model. The invention relates to the field of computer vision. A knowledge migration network based on vein image similarity is adopted to train and finely tune a face recognition model, a vein identity authentication model and a vein gender judgment model in sequence, and the finely tuned network performs feature extraction on a vein image. A supervision word bag model is adopted to carry out secondary coding on a high-dimensional gender feature vector output by a vein gender determination model, and identity authentication and gender determination are carried out. According to the knowledge migration network and the supervision word bag model based on the similar images, the similarity attributes between the neighborhood models can be utilized to enable the featurerepresentation parameter space of the model before fine adjustment to be crossed, the accuracy of identity recognition is improved, and the discrimination and generalization performance of the model is ensured.

Description

technical field [0001] The invention relates to the field of hand vein recognition, in particular to a multi-modal identity authentication method based on a knowledge transfer network of similar vein images. Background technique [0002] Venous blood vessels are one of the most important structures for humans to carry nutrients and metabolites, and compared with other biometric features (such as fingerprints, iris, gestures and faces), they have the characteristics of anti-counterfeiting and easy acceptance, and have become the most popular One of the methods of personal identification. Besides, the high convenience of image acquisition and robust feature representation lead to a more extensive and accurate vein-based personal recognition system. [0003] Although there are potential advantages in designing a robust identity authentication system based on vein recognition technology, in traditional feature extraction methods, the source vein image training library is small ...

AI Technical Summary

Problems solved by technology

However, the above feature encoding mode has flaws in its adaptability to problem samples such as rotation.

Images