Yes, FaceMe SDK can function offline without requiring an Internet connection.

After signing a Non-Disclosure Agreement (NDA), we can provide a trial version for your evaluation. Please contact us to request a trial version:

Contact Us for Trial Version

FaceMe SDK is compatible with a variety of low-power CPUs, including:

1. Qualcomm Snapdragon 410 and QCS603
2. Intel Celeron series
3. MediaTek i350
4. Rockchip RK3288
5. Broadcom BCM2711 (Raspberry Pi)
6. NVIDIA Jetson Nano

If you are uncertain whether your system meets the minimum requirements, please contact us for further assistance.

Supported Platforms and Hardware for FaceMe SDK:

Supported Operating Systems:

1. Windows
2. Ubuntu x64
3. Ubuntu ARM (64-bit)
4. Android
5. iOS
6. Debian
7. Jetson JetPack
8. Yocto
9. RedHat

Supported CPUs and SoCs:

1. Intel & AMD x86
2. Qualcomm Snapdragon series and QCS603
3. NVIDIA Jetson series
4. MediaTek i350
5. Rockchip
6. NXP
7. Broadcom chips

Supported GPUs/VPUs:

1. NVIDIA GPUs
2. Intel Movidius Myriad X

For more specific details or if you have additional questions about compatibility, please contact us.

"

FaceMe SDK supports the following operating systems and hardware:

1. Ubuntu x64
2. Ubuntu ARM64
3. Jetson
4. Red Hat

Please get in touch with us for further information and support if your hardware or operating system is not listed above.

"

Yes, the FaceMe SDK for Windows includes a Python SDK. You can find it in the following directory:

'%InstallPath%\FaceMeSDK\SDK\Other Languages\Python'

Face Extraction Models in FaceMe SDK:

1. Ultra High Precision Model (UH):

   • Features: Requires NVIDIA GPU or Movidius, providing the highest accuracy.
   • Recommended For: Servers, workstations, or high-performance personal computers.

2. Very High Precision Model (VH):

   • Features: Computational demand is only one-tenth of the UH model but with slightly lower accuracy.
   • Recommended For: Advanced mobile devices (Snapdragon 600 or higher) and common personal computers (Celeron or higher).

3. High Precision Model (H):

   • Features: Designed for mobile devices and streamlined users, suitable for face front verification scenarios.
   • Recommended For: Mobile devices and low-power systems.

If your platform or usage scenario is not covered by the above models, please contact us for further information.

The minFaceWidthRatio determines the minimum size of a face that the SDK will search for in an input image. This setting is a ratio relative to the width of the input image.

Setting the Ratio:

1. Example:
   • If the minFaceWidthRatio is set to 0.1, the smallest detectable face will be approximately 10% of the image width.
2. Range:
   • The ratio can range from 0.025 to 1, with a commonly recommended value of 0.05.

Important Considerations:

1. The minimal face ratio should not be less than 0.025 of the larger dimension of the image.
2. The face width should not be smaller than 32 pixels.

Adjusting the minFaceWidthRatio allows you to balance between detecting smaller faces and maintaining accuracy based on your specific application needs