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 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 2.45, ONNX Runtime 1.25.0 | Download |
| ONNX | w8a8 | Universal | QAIRT 2.45, ONNX Runtime 1.25.0 | Download |
| QNN_DLC | float | Universal | QAIRT 2.45 | Download |
| QNN_DLC | w8a8 | Universal | QAIRT 2.45 | Download |
| TFLITE | float | Universal | QAIRT 2.45 | Download |
| TFLITE | w8a8 | Universal | QAIRT 2.45 | Download |
For more device-specific assets and performance metrics, visit Facial-Landmark-Detection 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 Facial-Landmark-Detection on GitHub 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® X2 Elite | 0.184 ms | 213 - 213 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Snapdragon® X Elite | 0.265 ms | 181 - 181 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.244 ms | 0 - 19 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 0.538 ms | 0 - 25 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.319 ms | 0 - 19 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Qualcomm® QCS8450 | 0.538 ms | 0 - 25 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Snapdragon® 8 Elite Mobile | 0.218 ms | 0 - 24 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.206 ms | 0 - 16 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Qualcomm® QCS9075 | 0.446 ms | 0 - 45 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Qualcomm® QCS8750 | 0.218 ms | 0 - 24 MB | NPU |
| Facial-Landmark-Detection | ONNX | float | Qualcomm® QCS7181 | 0.265 ms | 181 - 181 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® X2 Elite | 0.077 ms | 213 - 213 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® X Elite | 0.124 ms | 149 - 149 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.132 ms | 0 - 28 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 8 Gen 1 Mobile | 0.263 ms | 0 - 35 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS6490 | 0.816 ms | 0 - 51 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.17 ms | 0 - 10 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS8450 | 0.263 ms | 0 - 35 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.119 ms | 0 - 18 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.223 ms | 0 - 22 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCM6690 | 0.617 ms | 0 - 21 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS9075 | 0.264 ms | 0 - 51 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Snapdragon® 8 Elite Mobile | 0.122 ms | 0 - 22 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS7790 | 0.223 ms | 0 - 22 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS8750 | 0.122 ms | 0 - 22 MB | NPU |
| Facial-Landmark-Detection | ONNX | w8a8 | Qualcomm® QCS7181 | 0.124 ms | 149 - 149 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® X2 Elite | 0.235 ms | 0 - 0 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® X Elite | 0.356 ms | 0 - 0 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.236 ms | 0 - 29 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 0.544 ms | 0 - 34 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS8275 | 1.153 ms | 0 - 21 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 0.299 ms | 0 - 2 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS8450 | 0.544 ms | 0 - 34 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 0.206 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® SA8295P | 0.696 ms | 0 - 17 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.196 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® SA7255P | 1.153 ms | 0 - 21 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS9075 | 0.391 ms | 0 - 2 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS8750 | 0.206 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | float | Qualcomm® QCS7181 | 0.356 ms | 0 - 0 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 0.164 ms | 0 - 0 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® X Elite | 0.223 ms | 1 - 1 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.137 ms | 0 - 36 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 8 Gen 1 Mobile | 0.237 ms | 0 - 43 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS6490 | 0.695 ms | 2 - 4 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS8275 | 0.456 ms | 0 - 21 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.162 ms | 0 - 1 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS8450 | 0.237 ms | 0 - 43 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.121 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.217 ms | 0 - 24 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCM6690 | 0.603 ms | 0 - 24 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS9075 | 0.223 ms | 2 - 4 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® SA7255P | 0.456 ms | 0 - 21 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Snapdragon® 8 Elite Mobile | 0.116 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® SA8295P | 0.468 ms | 0 - 18 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS7790 | 0.217 ms | 0 - 24 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS8750 | 0.116 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | QNN_DLC | w8a8 | Qualcomm® QCS7181 | 0.223 ms | 1 - 1 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.22 ms | 0 - 39 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 0.479 ms | 0 - 46 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® QCS8275 | 1.14 ms | 0 - 22 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 0.274 ms | 0 - 8 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA8775P | 3.615 ms | 0 - 29 MB | GPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA8650P | 3.615 ms | 0 - 29 MB | GPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA8255P | 3.615 ms | 0 - 29 MB | GPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® QCS8450 | 0.479 ms | 0 - 46 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Snapdragon® 8 Elite Mobile | 0.206 ms | 0 - 21 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA8295P | 0.672 ms | 0 - 18 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.193 ms | 0 - 21 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® SA7255P | 1.14 ms | 0 - 22 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® QCS9075 | 0.384 ms | 0 - 13 MB | NPU |
| Facial-Landmark-Detection | TFLITE | float | Qualcomm® QCS8750 | 0.206 ms | 0 - 21 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.14 ms | 0 - 36 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 8 Gen 1 Mobile | 0.237 ms | 0 - 41 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS6490 | 0.617 ms | 0 - 7 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS8275 | 0.487 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS8550 (Proxy) | 0.172 ms | 0 - 8 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA8775P | 3.708 ms | 0 - 31 MB | GPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA8650P | 3.708 ms | 0 - 31 MB | GPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA8255P | 3.708 ms | 0 - 31 MB | GPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS8450 | 0.237 ms | 0 - 41 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.12 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.22 ms | 0 - 23 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCM6690 | 0.6 ms | 0 - 23 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS9075 | 0.245 ms | 0 - 7 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA7255P | 0.487 ms | 0 - 20 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Snapdragon® 8 Elite Mobile | 0.117 ms | 0 - 24 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® SA8295P | 0.479 ms | 0 - 17 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS7790 | 0.22 ms | 0 - 23 MB | NPU |
| Facial-Landmark-Detection | TFLITE | w8a8 | Qualcomm® QCS8750 | 0.117 ms | 0 - 24 MB | NPU |
License
- The license for the original implementation of Facial-Landmark-Detection can be found here.
Community
- Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
- For questions or feedback please reach out to us.
