File size: 11,854 Bytes

---
library_name: pytorch
license: bsd-3-clause
pipeline_tag: object-detection
tags:
- real_time
- quantized
- android

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/face_det_lite_quantized/web-assets/model_demo.png)

# Lightweight-Face-Detection-Quantized: Optimized for Mobile Deployment
## face_det_lite_quantized is a face detection model


face_det_lite_quantized is a machine learning model that detect face in the images

This model is an implementation of Lightweight-Face-Detection-Quantized found [here](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/face_det_lite/model.py).


This repository provides scripts to run Lightweight-Face-Detection-Quantized on Qualcomm® devices.
More details on model performance across various devices, can be found
[here](https://aihub.qualcomm.com/models/face_det_lite_quantized).


### Model Details

- **Model Type:** Object detection
- **Model Stats:**
  - Model checkpoint: qfd360_sl_model.pt
  - Inference latency: RealTime
  - Input resolution: 480x640

| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|---|---|---|---|---|---|---|---|---|
| Lightweight-Face-Detection-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 0.276 ms | 0 - 9 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.341 ms | 0 - 9 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.so](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.so) |
| Lightweight-Face-Detection-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 0.631 ms | 0 - 8 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.onnx](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.onnx) |
| Lightweight-Face-Detection-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.201 ms | 0 - 26 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.244 ms | 0 - 26 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.so](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.so) |
| Lightweight-Face-Detection-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 0.448 ms | 0 - 28 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.onnx](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.onnx) |
| Lightweight-Face-Detection-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.201 ms | 0 - 19 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.265 ms | 0 - 19 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 0.407 ms | 0 - 21 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.onnx](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.onnx) |
| Lightweight-Face-Detection-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 0.798 ms | 0 - 17 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 1.039 ms | 0 - 11 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 6.608 ms | 0 - 3 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 0.278 ms | 0 - 10 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.339 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | SA7255P ADP | SA7255P | TFLITE | 3.129 ms | 0 - 13 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | SA7255P ADP | SA7255P | QNN | 3.463 ms | 0 - 9 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 0.28 ms | 0 - 9 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.363 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | SA8295P ADP | SA8295P | TFLITE | 0.666 ms | 0 - 19 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | SA8295P ADP | SA8295P | QNN | 0.994 ms | 0 - 14 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 0.278 ms | 0 - 9 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.336 ms | 0 - 2 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | SA8775P ADP | SA8775P | TFLITE | 0.587 ms | 0 - 14 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | SA8775P ADP | SA8775P | QNN | 0.814 ms | 0 - 10 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 0.366 ms | 0 - 20 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.tflite](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.tflite) |
| Lightweight-Face-Detection-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 0.455 ms | 0 - 24 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.465 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
| Lightweight-Face-Detection-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 0.702 ms | 3 - 3 MB | INT8 | NPU | [Lightweight-Face-Detection-Quantized.onnx](https://huggingface.co/qualcomm/Lightweight-Face-Detection-Quantized/blob/main/Lightweight-Face-Detection-Quantized.onnx) |




## Installation

This model can be installed as a Python package via pip.

```bash
pip install qai-hub-models
```


## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device

Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.

With this API token, you can configure your client to run models on the cloud
hosted devices.
```bash
qai-hub configure --api_token API_TOKEN
```
Navigate to [docs](https://app.aihub.qualcomm.com/docs/) for more information.



## Demo off target

The package contains a simple end-to-end demo that downloads pre-trained
weights and runs this model on a sample input.

```bash
python -m qai_hub_models.models.face_det_lite_quantized.demo
```

The above demo runs a reference implementation of pre-processing, model
inference, and post processing.

**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
environment, please add the following to your cell (instead of the above).
```
%run -m qai_hub_models.models.face_det_lite_quantized.demo
```


### Run model on a cloud-hosted device

In addition to the demo, you can also run the model on a cloud-hosted Qualcomm®
device. This script does the following:
* Performance check on-device on a cloud-hosted device
* Downloads compiled assets that can be deployed on-device for Android.
* Accuracy check between PyTorch and on-device outputs.

```bash
python -m qai_hub_models.models.face_det_lite_quantized.export
```
```
Profiling Results
------------------------------------------------------------
Lightweight-Face-Detection-Quantized
Device                          : Samsung Galaxy S23 (13)
Runtime                         : TFLITE                 
Estimated inference time (ms)   : 0.3                    
Estimated peak memory usage (MB): [0, 9]                 
Total # Ops                     : 90                     
Compute Unit(s)                 : NPU (90 ops)           
```




## Run demo on a cloud-hosted device

You can also run the demo on-device.

```bash
python -m qai_hub_models.models.face_det_lite_quantized.demo --on-device
```

**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
environment, please add the following to your cell (instead of the above).
```
%run -m qai_hub_models.models.face_det_lite_quantized.demo -- --on-device
```


## Deploying compiled model to Android


The models can be deployed using multiple runtimes:
- TensorFlow Lite (`.tflite` export): [This
  tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a
  guide to deploy the .tflite model in an Android application.


- QNN (`.so` export ): This [sample
  app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html)
provides instructions on how to use the `.so` shared library  in an Android application.


## View on Qualcomm® AI Hub
Get more details on Lightweight-Face-Detection-Quantized's performance across various devices [here](https://aihub.qualcomm.com/models/face_det_lite_quantized).
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)


## License
* The license for the original implementation of Lightweight-Face-Detection-Quantized can be found [here](https://github.com/qcom-ai-hub/ai-hub-models-internal/blob/main/LICENSE).
* The license for the compiled assets for on-device deployment can be found [here](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf)



## References
* [None](None)
* [Source Model Implementation](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/face_det_lite/model.py)



## Community
* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
* For questions or feedback please [reach out to us](mailto:[email protected]).