A novel approach to generate cancellable biometric templates using style transfer and hash codes

Key Points

• Cancellable biometric templates can enhance data security with unique hash codes, offering improved privacy.
• The method utilizes style transfer techniques to create variations of biometric data for secure identification.
• Using a combination of deep learning and hashing, the approach developed novel templates for biometric systems.
• These advancements indicate a potential shift towards more secure methods in biometric data handling.