Hugging Face's logo

qualcomm
/
MobileFaceNet

Object Detection
PyTorch
real_time
bu_auto
android
Model card Files
xet
 Community

MobileFaceNet: Optimized for Qualcomm Devices

MobileFaceNet is an efficient CNN that maps a 112x112 face image to a compact 128-dimensional embedding. Two embeddings are compared via cosine similarity to determine whether they belong to the same person, achieving 99.48% accuracy on the LFW benchmark. The model uses depthwise-separable convolutions and inverted residual blocks (MobileNetV2-style) to stay under 1M parameters, making it well-suited for real-time face verification on mobile and edge devices. Trained with ArcFace loss on MS-Celeb-1M.

This is based on the implementation of MobileFaceNet found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up to run these models on a hosted Qualcomm® device.

Getting Started

There are two ways to deploy this model on your device:

Option 1: Download Pre-Exported Models

Below are pre-exported model assets ready for deployment.

Runtime Precision Chipset SDK Versions Download
ONNX float Universal QAIRT 2.50, ONNX Runtime 1.27.1 Download
ONNX w8a16 Universal QAIRT 2.50, ONNX Runtime 1.27.1 Download
QNN_DLC float Universal QAIRT 2.50 Download
QNN_DLC w8a16 Universal QAIRT 2.50 Download
TFLITE float Universal QAIRT 2.50 Download

For more device-specific assets and performance metrics, visit MobileFaceNet on Qualcomm® AI Hub.

Option 2: Export with Custom Configurations

Use the Qualcomm® AI Hub Models Python library to compile and export the model with your own:

  • Custom weights (e.g., fine-tuned checkpoints)
  • Custom input shapes
  • Target device and runtime configurations

This option is ideal if you need to customize the model beyond the default configuration provided here.

See our repository for MobileFaceNet on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.object_detection

Model Stats:

  • Embedding dimension: 128
  • Input resolution: 112x112
  • Model checkpoint: mobilefacenet.pt
  • Model size (float): 4MB
  • Number of parameters: 1M

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
MobileFaceNet ONNX float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.531 ms 0 - 38 MB NPU
MobileFaceNet ONNX float Snapdragon® 8 Elite For Galaxy Mobile 0.545 ms 0 - 41 MB NPU
MobileFaceNet ONNX float Snapdragon® X2 Elite 0.609 ms 1 - 1 MB NPU
MobileFaceNet ONNX float Snapdragon® X Elite 1.034 ms 0 - 0 MB NPU
MobileFaceNet ONNX float Snapdragon® 8 Gen 3 Mobile 0.699 ms 0 - 57 MB NPU
MobileFaceNet ONNX float Snapdragon® 8 Gen 1 Mobile 1.467 ms 0 - 59 MB NPU
MobileFaceNet ONNX float Qualcomm® Dragonwing™ IQ-8275 1.457 ms 0 - 4 MB NPU
MobileFaceNet ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 1.001 ms 0 - 5 MB NPU
MobileFaceNet ONNX float Qualcomm® QCS8450 1.467 ms 0 - 59 MB NPU
MobileFaceNet ONNX float Qualcomm® Dragonwing™ IQ-9075 1.393 ms 0 - 4 MB NPU
MobileFaceNet ONNX float Qualcomm® Dragonwing™ IQ-X7181 1.034 ms 0 - 0 MB NPU
MobileFaceNet ONNX float Qualcomm® Dragonwing™ Q-8750 0.545 ms 0 - 41 MB NPU
MobileFaceNet ONNX w8a16 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.377 ms 0 - 47 MB NPU
MobileFaceNet ONNX w8a16 Snapdragon® 8 Elite For Galaxy Mobile 0.451 ms 0 - 51 MB NPU
MobileFaceNet ONNX w8a16 Snapdragon® X2 Elite 0.459 ms 1 - 1 MB NPU
MobileFaceNet ONNX w8a16 Snapdragon® X Elite 0.819 ms 0 - 0 MB NPU
MobileFaceNet ONNX w8a16 Snapdragon® 8 Gen 3 Mobile 0.575 ms 0 - 59 MB NPU
MobileFaceNet ONNX w8a16 Snapdragon® 8 Gen 1 Mobile 1.054 ms 0 - 61 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® Dragonwing™ QCS6490 2.372 ms 0 - 3 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® Dragonwing™ IQ-8275 0.879 ms 0 - 4 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 0.797 ms 0 - 3 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® QCS8450 1.054 ms 0 - 61 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® Dragonwing™ IQ-9075 0.907 ms 0 - 3 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® Dragonwing™ IQ-X7181 0.819 ms 0 - 0 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® Dragonwing™ Q-6690 5.437 ms 0 - 158 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® Dragonwing™ Q-7790 0.924 ms 0 - 46 MB NPU
MobileFaceNet ONNX w8a16 Qualcomm® Dragonwing™ Q-8750 0.451 ms 0 - 51 MB NPU
MobileFaceNet ONNX w8a16 Snapdragon® 7 Gen 4 Mobile 0.924 ms 0 - 46 MB NPU
MobileFaceNet QNN_DLC float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.532 ms 0 - 35 MB NPU
MobileFaceNet QNN_DLC float Snapdragon® 8 Elite For Galaxy Mobile 0.591 ms 0 - 34 MB NPU
MobileFaceNet QNN_DLC float Snapdragon® X2 Elite 0.733 ms 0 - 0 MB NPU
MobileFaceNet QNN_DLC float Snapdragon® X Elite 1.343 ms 0 - 0 MB NPU
MobileFaceNet QNN_DLC float Snapdragon® 8 Gen 3 Mobile 0.786 ms 0 - 50 MB NPU
MobileFaceNet QNN_DLC float Snapdragon® 8 Gen 1 Mobile 1.624 ms 0 - 55 MB NPU
MobileFaceNet QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 1.464 ms 0 - 3 MB NPU
MobileFaceNet QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 1.171 ms 0 - 30 MB NPU
MobileFaceNet QNN_DLC float Qualcomm® QCS8450 1.624 ms 0 - 55 MB NPU
MobileFaceNet QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 1.496 ms 0 - 3 MB NPU
MobileFaceNet QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 1.343 ms 0 - 0 MB NPU
MobileFaceNet QNN_DLC float Qualcomm® Dragonwing™ Q-8750 0.591 ms 0 - 34 MB NPU
MobileFaceNet QNN_DLC float Qualcomm® SA8295P 1.907 ms 0 - 33 MB NPU
MobileFaceNet QNN_DLC w8a16 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.389 ms 0 - 42 MB NPU
MobileFaceNet QNN_DLC w8a16 Snapdragon® 8 Elite For Galaxy Mobile 0.497 ms 0 - 41 MB NPU
MobileFaceNet QNN_DLC w8a16 Snapdragon® X2 Elite 0.585 ms 0 - 0 MB NPU
MobileFaceNet QNN_DLC w8a16 Snapdragon® X Elite 1.154 ms 0 - 0 MB NPU
MobileFaceNet QNN_DLC w8a16 Snapdragon® 8 Gen 3 Mobile 0.682 ms 0 - 53 MB NPU
MobileFaceNet QNN_DLC w8a16 Snapdragon® 8 Gen 1 Mobile 1.202 ms 0 - 58 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® Dragonwing™ QCS6490 2.705 ms 0 - 2 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-8275 0.966 ms 0 - 3 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 0.997 ms 0 - 2 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® SA8650P 1.205 ms 0 - 38 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® SA8255P 1.205 ms 0 - 38 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® QCS8450 1.202 ms 0 - 58 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-9075 1.091 ms 0 - 2 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-X7181 1.154 ms 0 - 0 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-6690 6.035 ms 0 - 148 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-7790 1.081 ms 0 - 40 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-8750 0.497 ms 0 - 41 MB NPU
MobileFaceNet QNN_DLC w8a16 Qualcomm® SA8295P 1.456 ms 0 - 40 MB NPU
MobileFaceNet QNN_DLC w8a16 Snapdragon® 7 Gen 4 Mobile 1.081 ms 0 - 40 MB NPU
MobileFaceNet TFLITE float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 0.533 ms 0 - 36 MB NPU
MobileFaceNet TFLITE float Snapdragon® 8 Elite For Galaxy Mobile 0.562 ms 0 - 35 MB NPU
MobileFaceNet TFLITE float Snapdragon® 8 Gen 3 Mobile 0.699 ms 9 - 60 MB NPU
MobileFaceNet TFLITE float Snapdragon® 8 Gen 1 Mobile 1.411 ms 0 - 53 MB NPU
MobileFaceNet TFLITE float Qualcomm® Dragonwing™ IQ-8275 1.442 ms 0 - 6 MB NPU
MobileFaceNet TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 0.976 ms 0 - 2 MB NPU
MobileFaceNet TFLITE float Qualcomm® SA8775P 16.232 ms 1 - 6 MB CPU
MobileFaceNet TFLITE float Qualcomm® SA8650P 16.232 ms 1 - 6 MB CPU
MobileFaceNet TFLITE float Qualcomm® SA8255P 16.232 ms 1 - 6 MB CPU
MobileFaceNet TFLITE float Qualcomm® QCS8450 1.411 ms 0 - 53 MB NPU
MobileFaceNet TFLITE float Qualcomm® Dragonwing™ IQ-9075 1.357 ms 0 - 5 MB NPU
MobileFaceNet TFLITE float Qualcomm® Dragonwing™ Q-8750 0.562 ms 0 - 35 MB NPU
MobileFaceNet TFLITE float Qualcomm® SA7255P 4.497 ms 0 - 38 MB NPU
MobileFaceNet TFLITE float Qualcomm® SA8295P 1.702 ms 0 - 33 MB NPU

License

  • The license for the original implementation of MobileFaceNet can be found here.

References

Community

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Object Detection
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support