FaceAISDK/FaceAISDK_Android

What it solves

FaceAISDK provides an offline, on-device face recognition system for Android devices. It eliminates the need for cloud-based processing, which reduces latency, removes network dependency and ensures user privacy by keeping all biometric data on the local device.

How it works

The SDK implements edge-side inference for face detection, recognition, and liveness detection. It supports both standard system cameras and UVC protocol USB cameras. The system can perform 1:1 verification (comparing a person to a specific identity) and 1:N search (identifying a person from a library of faces).

Who it’s for

Developers building Android applications for secure access control, password-less login, mobile attendance, and smart home or robotics integrations where offline functionality and data privacy are critical.

Highlights

• Fully Offline: All functions work without a network connection, ensuring no data is uploaded to the cloud.
• Liveness Detection: Supports both silent liveness detection and action-based checks (blinking, smiling, head nodding/shaking).
• Flexible Search: Capable of 1:1 face verification and 1:N/M:N face search and recognition.
• Hardware Support: Compatible with Android 8 through 16, supporting both built-in phone cameras and external USB cameras via UVC protocol.