Cloud-Based Unified Credential Storage: Enhancing Security Across Multiple Accounts

The Unified Secure Cloud Storage for Multi-Account Credentials successfully integrates advanced facial recognition technology with secure cloud-based storage to enhance the protection of sensitive user data. The Android application securely stores multi-account credentials using Encrypted Shared Preferences, ensuring local encryption and restricted access. During registration, ten facial images of the user are captured and uploaded to an AWS S3 bucket via a Django backend, enabling future identity verification. Authentication is performed by capturing a login image, which is processed using the Deep Face VGG Face model to compare against stored images. The system demonstrated 92% facial recognition accuracy, effectively minimizing false positives by requiring a match in at least six out of ten images. AWS integration enabled low-latency verification with an average response time of 2- 3 seconds, ensuring a smooth user experience. Testing with diverse users yielded a 90% success rate in authentication, reinforcing the system’s reliability. Users appreciated the convenience and security of facial recognition over traditional password-based systems. This project lays the foundation for future advancements such as liveness detection and spoofing prevention, contributing to the evolution of secure mobile authentication solutions.