README.md

---
library_name: pytorch
license: other
tags:
- backbone
- android
pipeline_tag: keypoint-detection

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/facemap_3dmm/web-assets/model_demo.png)

# Facial-Landmark-Detection: Optimized for Qualcomm Devices

Detects facial landmarks (eg, nose, mouth, etc.). This model's architecture was developed by Qualcomm. The model was trained by Qualcomm on a proprietary dataset of faces, but can be used on any image.

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/blob/main/src/qai_hub_models/models/facemap_3dmm) 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.com) 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 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/facemap_3dmm/releases/v0.54.0/facemap_3dmm-onnx-float.zip)
| ONNX | w8a8 | Universal | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/facemap_3dmm/releases/v0.54.0/facemap_3dmm-onnx-w8a8.zip)
| QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/facemap_3dmm/releases/v0.54.0/facemap_3dmm-qnn_dlc-float.zip)
| QNN_DLC | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/facemap_3dmm/releases/v0.54.0/facemap_3dmm-qnn_dlc-w8a8.zip)
| TFLITE | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/facemap_3dmm/releases/v0.54.0/facemap_3dmm-tflite-float.zip)
| TFLITE | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/facemap_3dmm/releases/v0.54.0/facemap_3dmm-tflite-w8a8.zip)

For more device-specific assets and performance metrics, visit **[Facial-Landmark-Detection on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/facemap_3dmm)**.


### Option 2: Export with Custom Configurations

Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/facemap_3dmm) 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 [Facial-Landmark-Detection on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/facemap_3dmm) for usage instructions.

## Model Details

**Model Type:** Model_use_case.pose_estimation

**Model Stats:**
- Input resolution: 128x128
- Number of parameters: 5.42M
- Model size (float): 20.7 MB
- Model size (w8a8): 5.27 MB

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| Facial-Landmark-Detection | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.211 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | ONNX | float | Snapdragon® 8 Elite Mobile | 0.227 ms | 0 - 18 MB | NPU
| Facial-Landmark-Detection | ONNX | float | Snapdragon® X2 Elite | 0.164 ms | 10 - 10 MB | NPU
| Facial-Landmark-Detection | ONNX | float | Snapdragon® X Elite | 0.396 ms | 10 - 10 MB | NPU
| Facial-Landmark-Detection | ONNX | float | Snapdragon® X Elite | 0.396 ms | 10 - 10 MB | NPU
| Facial-Landmark-Detection | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.239 ms | 0 - 33 MB | NPU
| Facial-Landmark-Detection | ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.32 ms | 0 - 25 MB | NPU
| Facial-Landmark-Detection | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.227 ms | 0 - 18 MB | NPU
| Facial-Landmark-Detection | ONNX | float | Qualcomm® QCS9075 | 0.439 ms | 0 - 3 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.112 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 8 Elite Mobile | 0.119 ms | 0 - 20 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® X2 Elite | 0.076 ms | 5 - 5 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® X Elite | 0.236 ms | 5 - 5 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® X Elite | 0.236 ms | 5 - 5 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.141 ms | 0 - 38 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS6490 | 3.343 ms | 4 - 11 MB | CPU
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.167 ms | 0 - 13 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 1.397 ms | 0 - 7 MB | CPU
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCM6690 | 1.846 ms | 0 - 6 MB | CPU
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS9075 | 0.293 ms | 0 - 3 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 0.119 ms | 0 - 20 MB | NPU
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 1.397 ms | 0 - 7 MB | CPU
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.189 ms | 0 - 24 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 0.21 ms | 0 - 20 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® X2 Elite | 0.237 ms | 0 - 0 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® X Elite | 0.373 ms | 0 - 0 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® X Elite | 0.373 ms | 0 - 0 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.237 ms | 0 - 30 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 0.305 ms | 0 - 2 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® SA8775P | 0.497 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® SA8775P | 0.497 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® SA8775P | 0.497 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® SA7255P | 1.15 ms | 0 - 21 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 0.545 ms | 0 - 30 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® SA8295P | 0.674 ms | 0 - 18 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.21 ms | 0 - 20 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS9075 | 0.392 ms | 0 - 2 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.111 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 8 Elite Mobile | 0.114 ms | 0 - 20 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 0.184 ms | 0 - 0 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® X Elite | 0.239 ms | 0 - 0 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® X Elite | 0.239 ms | 0 - 0 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.135 ms | 0 - 36 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS6490 | 0.685 ms | 0 - 2 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.167 ms | 0 - 58 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® SA8775P | 0.317 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® SA8775P | 0.317 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® SA8775P | 0.317 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.218 ms | 0 - 24 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCM6690 | 0.589 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS9075 | 0.223 ms | 0 - 2 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 0.114 ms | 0 - 20 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS8450 (Proxy) | 0.237 ms | 0 - 38 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® SA7255P | 0.451 ms | 0 - 21 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® SA8295P | 0.454 ms | 0 - 19 MB | NPU
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.218 ms | 0 - 24 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.185 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Snapdragon® 8 Elite Mobile | 0.209 ms | 0 - 26 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.22 ms | 0 - 40 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 0.275 ms | 0 - 102 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA8775P | 0.499 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA8775P | 0.499 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA8775P | 0.499 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA7255P | 1.145 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 0.483 ms | 0 - 41 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA8295P | 0.679 ms | 0 - 18 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.209 ms | 0 - 26 MB | NPU
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® QCS9075 | 0.386 ms | 0 - 13 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.109 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 8 Elite Mobile | 0.124 ms | 0 - 20 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.147 ms | 0 - 36 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS6490 | 0.564 ms | 0 - 7 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.173 ms | 0 - 1 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA8775P | 0.332 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA8775P | 0.332 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA8775P | 0.332 ms | 0 - 22 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.218 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCM6690 | 0.604 ms | 0 - 23 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS9075 | 0.235 ms | 0 - 7 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 0.124 ms | 0 - 20 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS8450 (Proxy) | 0.236 ms | 0 - 37 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA7255P | 0.467 ms | 0 - 21 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA8295P | 0.47 ms | 0 - 18 MB | NPU
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.218 ms | 0 - 23 MB | NPU

## License
* The license for the original implementation of Facial-Landmark-Detection can be found
  [here](https://github.com/qualcomm/ai-hub-models/blob/main/LICENSE).



## 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:ai-hub-support@qti.qualcomm.com).