---
title: HRNetFace
canonical_url: "https://www.modelscope.cn/models/Qualcomm-AI-Hub/HRNetFace"
md_url: "https://www.modelscope.cn/models/Qualcomm-AI-Hub/HRNetFace.md"
repository: Qualcomm-AI-Hub/HRNetFace
last_updated: 2026-10-16
license: mit
downloads: 0
stars: 0
tags:
  - real_time
  - android
---

# HRNetFace

> HRNetFace - Qualcomm-AI-Hub 在 ModelScope 开源的模型。HRNetFace: Optimized for Qualcomm Devices

Qualcomm-AI-Hub/HRNetFace 是 ModelScope 魔搭社区上的机器学习模型，采用 mit 许可。

- **Repository**: Qualcomm-AI-Hub/HRNetFace
- **License**: mit
- **Tags**: real_time, android
- **Downloads**: 0
- **Stars**: 0
- **Last updated**: 2026-10-16

Source: https://www.modelscope.cn/models/Qualcomm-AI-Hub/HRNetFace

---

![](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/hrnet_face/web-assets/model_demo.png)

# HRNetFace: Optimized for Qualcomm Devices

Detects attributes (liveness, eye closeness, mask presence, glasses presence, sunglasses presence) that apply to a given face.

This is based on the implementation of HRNetFace found [here](https://github.com/HRNet/HRNet-Facial-Landmark-Detection).
This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models-cn/blob/main/src/qai_hub_models/models/hrnet_face) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).

Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.cn) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) 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 2.45, ONNX Runtime 1.27.1 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/hrnet_face/releases/v0.62.1/hrnet_face-onnx-float.zip)
| ONNX | w8a8 | Universal | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/hrnet_face/releases/v0.62.1/hrnet_face-onnx-w8a8.zip)
| QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/hrnet_face/releases/v0.62.1/hrnet_face-qnn_dlc-float.zip)
| QNN_DLC | w8a8 | Universal | QAIRT 2.45 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/hrnet_face/releases/v0.62.1/hrnet_face-qnn_dlc-w8a8.zip)
| TFLITE | float | Universal | QAIRT 2.45 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/hrnet_face/releases/v0.62.1/hrnet_face-tflite-float.zip)
| TFLITE | w8a8 | Universal | QAIRT 2.45 | [Download](https://china-qaihub-public-assets.s3.cn-northwest-1.amazonaws.com.cn/qai-hub-models/models/hrnet_face/releases/v0.62.1/hrnet_face-tflite-w8a8.zip)

For more device-specific assets and performance metrics, visit **[HRNetFace on Qualcomm® AI Hub](https://aihub.qualcomm.cn/models/hrnet_face)**.


### Option 2: Export with Custom Configurations

Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models-cn/blob/main/src/qai_hub_models/models/hrnet_face) 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 [HRNetFace on GitHub](https://github.com/qualcomm/ai-hub-models-cn/blob/main/src/qai_hub_models/models/hrnet_face) for usage instructions.

## Model Details

**Model Type:** Model_use_case.object_detection

**Model Stats:**
- Input resolution: 256x256
- Model checkpoint: HR18-COFW.pth
- Model size (float): 36.87MB
- Model size (w8a8): 17.7 MB
- Number of parameters: 9.68M

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| HRNetFace | ONNX | float | Snapdragon® X2 Elite | 1.548 ms | 2 - 2 MB | NPU
| HRNetFace | ONNX | float | Snapdragon® X Elite | 3.122 ms | 31 - 31 MB | NPU
| HRNetFace | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 2.148 ms | 0 - 151 MB | NPU
| HRNetFace | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 4.547 ms | 1 - 135 MB | NPU
| HRNetFace | ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | 4.881 ms | 2 - 5 MB | NPU
| HRNetFace | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 2.987 ms | 0 - 34 MB | NPU
| HRNetFace | ONNX | float | Qualcomm® QCS8450 | 4.547 ms | 1 - 135 MB | NPU
| HRNetFace | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 4.612 ms | 2 - 4 MB | NPU
| HRNetFace | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 3.122 ms | 31 - 31 MB | NPU
| HRNetFace | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 1.65 ms | 0 - 94 MB | NPU
| HRNetFace | ONNX | float | Snapdragon® 8 Elite Mobile | 1.65 ms | 0 - 94 MB | NPU
| HRNetFace | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 1.344 ms | 1 - 96 MB | NPU
| HRNetFace | ONNX | w8a8 | Snapdragon® X2 Elite | 0.664 ms | 1 - 1 MB | NPU
| HRNetFace | ONNX | w8a8 | Snapdragon® X Elite | 1.326 ms | 16 - 16 MB | NPU
| HRNetFace | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.877 ms | 0 - 155 MB | NPU
| HRNetFace | ONNX | w8a8 | Snapdragon® 8 Gen 1 Mobile | 1.748 ms | 0 - 159 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 5.093 ms | 0 - 3 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® Dragonwing™ IQ-8275 | 1.396 ms | 0 - 4 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.291 ms | 0 - 18 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® QCS8450 | 1.748 ms | 0 - 159 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 1.511 ms | 0 - 3 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® Dragonwing™ IQ-X7181 | 1.326 ms | 16 - 16 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® Dragonwing™ Q-6690 | 10.047 ms | 0 - 233 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® Dragonwing™ Q-7790 | 1.612 ms | 0 - 114 MB | NPU
| HRNetFace | ONNX | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.705 ms | 0 - 107 MB | NPU
| HRNetFace | ONNX | w8a8 | Snapdragon® 8 Elite Mobile | 0.705 ms | 0 - 107 MB | NPU
| HRNetFace | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.587 ms | 0 - 109 MB | NPU
| HRNetFace | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 1.612 ms | 0 - 114 MB | NPU
| HRNetFace | QNN_DLC | float | Snapdragon® X2 Elite | 1.945 ms | 1 - 1 MB | NPU
| HRNetFace | QNN_DLC | float | Snapdragon® X Elite | 3.742 ms | 1 - 1 MB | NPU
| HRNetFace | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 2.331 ms | 0 - 121 MB | NPU
| HRNetFace | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 5.484 ms | 0 - 104 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-8275 | 4.963 ms | 1 - 4 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 3.272 ms | 1 - 142 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® SA8775P | 5.105 ms | 1 - 79 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® SA8650P | 5.105 ms | 1 - 79 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® SA8255P | 5.105 ms | 1 - 79 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® QCS8450 | 5.484 ms | 0 - 104 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-9075 | 4.839 ms | 1 - 4 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-X7181 | 3.742 ms | 1 - 1 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® Dragonwing™ Q-8750 | 1.756 ms | 0 - 76 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® SA7255P | 15.434 ms | 1 - 75 MB | NPU
| HRNetFace | QNN_DLC | float | Qualcomm® SA8295P | 5.664 ms | 1 - 63 MB | NPU
| HRNetFace | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 1.756 ms | 0 - 76 MB | NPU
| HRNetFace | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 1.411 ms | 1 - 79 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 0.89 ms | 0 - 0 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Snapdragon® X Elite | 1.548 ms | 0 - 0 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.892 ms | 0 - 110 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Snapdragon® 8 Gen 1 Mobile | 1.858 ms | 0 - 109 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 5.73 ms | 2 - 4 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ IQ-8275 | 1.394 ms | 0 - 3 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.327 ms | 0 - 105 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® SA8775P | 1.694 ms | 0 - 72 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® SA8650P | 1.694 ms | 0 - 72 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® SA8255P | 1.694 ms | 0 - 72 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® QCS8450 | 1.858 ms | 0 - 109 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 1.533 ms | 0 - 2 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ IQ-X7181 | 1.548 ms | 0 - 0 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-6690 | 10.633 ms | 0 - 199 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-7790 | 1.623 ms | 0 - 79 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.687 ms | 0 - 73 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® SA7255P | 3.248 ms | 0 - 71 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Qualcomm® SA8295P | 2.201 ms | 0 - 69 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Snapdragon® 8 Elite Mobile | 0.687 ms | 0 - 73 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.545 ms | 0 - 74 MB | NPU
| HRNetFace | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 1.623 ms | 0 - 79 MB | NPU
| HRNetFace | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 2.335 ms | 0 - 137 MB | NPU
| HRNetFace | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 5.491 ms | 0 - 125 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 5.011 ms | 0 - 35 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 3.313 ms | 0 - 3 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® SA8775P | 5.144 ms | 0 - 91 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® SA8650P | 5.144 ms | 0 - 91 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® SA8255P | 5.144 ms | 0 - 91 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® QCS8450 | 5.491 ms | 0 - 125 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 4.875 ms | 0 - 35 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 1.757 ms | 0 - 96 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® SA7255P | 15.461 ms | 1 - 92 MB | NPU
| HRNetFace | TFLITE | float | Qualcomm® SA8295P | 5.655 ms | 1 - 74 MB | NPU
| HRNetFace | TFLITE | float | Snapdragon® 8 Elite Mobile | 1.757 ms | 0 - 96 MB | NPU
| HRNetFace | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 1.408 ms | 0 - 88 MB | NPU
| HRNetFace | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.749 ms | 0 - 114 MB | NPU
| HRNetFace | TFLITE | w8a8 | Snapdragon® 8 Gen 1 Mobile | 1.532 ms | 0 - 114 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 5.523 ms | 0 - 18 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® Dragonwing™ IQ-8275 | 1.202 ms | 0 - 18 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.117 ms | 0 - 3 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® SA8775P | 1.48 ms | 0 - 74 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® SA8650P | 1.48 ms | 0 - 74 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® SA8255P | 1.48 ms | 0 - 74 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® QCS8450 | 1.532 ms | 0 - 114 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 1.295 ms | 0 - 18 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-6690 | 9.834 ms | 0 - 193 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-7790 | 1.358 ms | 0 - 75 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.588 ms | 0 - 75 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® SA7255P | 2.885 ms | 0 - 71 MB | NPU
| HRNetFace | TFLITE | w8a8 | Qualcomm® SA8295P | 1.987 ms | 0 - 69 MB | NPU
| HRNetFace | TFLITE | w8a8 | Snapdragon® 8 Elite Mobile | 0.588 ms | 0 - 75 MB | NPU
| HRNetFace | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.472 ms | 0 - 73 MB | NPU
| HRNetFace | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 1.358 ms | 0 - 75 MB | NPU

## License
* The license for the original implementation of HRNetFace can be found
  [here](https://github.com/HRNet/HRNet-Facial-Landmark-Detection/blob/master/LICENCE).

## References
* [Deep High-Resolution Representation Learning for Visual Recognition](https://arxiv.org/abs/1908.07919)
* [Source Model Implementation](https://github.com/HRNet/HRNet-Facial-Landmark-Detection)

## Community
* For questions or feedback, chat with our engineers in [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.cn).