README.md

---
library_name: pytorch
license: apache-2.0
pipeline_tag: keypoint-detection
tags:
- real_time
- android

---

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

# MediaPipe-Pose-Estimation: Optimized for Mobile Deployment
## Detect and track human body poses in real-time images and video streams


The MediaPipe Pose Landmark Detector is a machine learning pipeline that predicts bounding boxes and pose skeletons of poses in an image.

This model is an implementation of MediaPipe-Pose-Estimation found [here](https://github.com/zmurez/MediaPipePyTorch/).


This repository provides scripts to run MediaPipe-Pose-Estimation on Qualcomm® devices.
More details on model performance across various devices, can be found
[here](https://aihub.qualcomm.com/models/mediapipe_pose).


### Model Details

- **Model Type:** Pose estimation
- **Model Stats:**
  - Input resolution: 256x256
  - Number of parameters (MediaPipePoseDetector): 815K
  - Model size (MediaPipePoseDetector): 3.14 MB
  - Number of parameters (MediaPipePoseLandmarkDetector): 3.37M
  - Model size (MediaPipePoseLandmarkDetector): 12.9 MB

| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|---|---|---|---|---|---|---|---|---|
| MediaPipePoseDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 0.78 ms | 0 - 11 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 1.01 ms | 0 - 4 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.onnx](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.onnx) |
| MediaPipePoseDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.56 ms | 0 - 46 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 0.734 ms | 0 - 49 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.onnx](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.onnx) |
| MediaPipePoseDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.563 ms | 0 - 24 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 0.757 ms | 0 - 26 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.onnx](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.onnx) |
| MediaPipePoseDetector | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 0.775 ms | 0 - 5 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 0.774 ms | 0 - 1 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 0.776 ms | 0 - 1 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 0.772 ms | 0 - 1 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | SA8295P ADP | SA8295P | TFLITE | 2.347 ms | 0 - 20 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 1.902 ms | 0 - 41 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.tflite) |
| MediaPipePoseDetector | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.062 ms | 3 - 3 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.onnx](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseDetector.onnx) |
| MediaPipePoseLandmarkDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 0.819 ms | 0 - 2 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 1.321 ms | 0 - 9 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.onnx](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.onnx) |
| MediaPipePoseLandmarkDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.615 ms | 0 - 90 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 1.0 ms | 0 - 95 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.onnx](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.onnx) |
| MediaPipePoseLandmarkDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.469 ms | 0 - 35 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 0.915 ms | 0 - 42 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.onnx](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.onnx) |
| MediaPipePoseLandmarkDetector | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 0.802 ms | 0 - 1 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 0.824 ms | 0 - 2 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 0.827 ms | 0 - 2 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 0.816 ms | 0 - 5 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | SA8295P ADP | SA8295P | TFLITE | 5.116 ms | 0 - 39 MB | FP16 | GPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 1.799 ms | 0 - 79 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.tflite](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.tflite) |
| MediaPipePoseLandmarkDetector | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.405 ms | 8 - 8 MB | FP16 | NPU | [MediaPipe-Pose-Estimation.onnx](https://huggingface.co/qualcomm/MediaPipe-Pose-Estimation/blob/main/MediaPipePoseLandmarkDetector.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.mediapipe_pose.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.mediapipe_pose.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.mediapipe_pose.export
```
```
Profiling Results
------------------------------------------------------------
MediaPipePoseDetector
Device                          : Samsung Galaxy S23 (13)
Runtime                         : TFLITE                 
Estimated inference time (ms)   : 0.8                    
Estimated peak memory usage (MB): [0, 11]                
Total # Ops                     : 106                    
Compute Unit(s)                 : NPU (106 ops)          

------------------------------------------------------------
MediaPipePoseLandmarkDetector
Device                          : Samsung Galaxy S23 (13)
Runtime                         : TFLITE                 
Estimated inference time (ms)   : 0.8                    
Estimated peak memory usage (MB): [0, 2]                 
Total # Ops                     : 219                    
Compute Unit(s)                 : NPU (219 ops)          
```


## How does this work?

This [export script](https://aihub.qualcomm.com/models/mediapipe_pose/qai_hub_models/models/MediaPipe-Pose-Estimation/export.py)
leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) to optimize, validate, and deploy this model
on-device. Lets go through each step below in detail:

Step 1: **Compile model for on-device deployment**

To compile a PyTorch model for on-device deployment, we first trace the model
in memory using the `jit.trace` and then call the `submit_compile_job` API.

```python
import torch

import qai_hub as hub
from qai_hub_models.models.mediapipe_pose import MediaPipePoseDetector,MediaPipePoseLandmarkDetector

# Load the model
pose_detector_model = MediaPipePoseDetector.from_pretrained()
pose_landmark_detector_model = MediaPipePoseLandmarkDetector.from_pretrained()

# Device
device = hub.Device("Samsung Galaxy S23")

# Trace model
pose_detector_input_shape = pose_detector_model.get_input_spec()
pose_detector_sample_inputs = pose_detector_model.sample_inputs()

traced_pose_detector_model = torch.jit.trace(pose_detector_model, [torch.tensor(data[0]) for _, data in pose_detector_sample_inputs.items()])

# Compile model on a specific device
pose_detector_compile_job = hub.submit_compile_job(
    model=traced_pose_detector_model ,
    device=device,
    input_specs=pose_detector_model.get_input_spec(),
)

# Get target model to run on-device
pose_detector_target_model = pose_detector_compile_job.get_target_model()
# Trace model
pose_landmark_detector_input_shape = pose_landmark_detector_model.get_input_spec()
pose_landmark_detector_sample_inputs = pose_landmark_detector_model.sample_inputs()

traced_pose_landmark_detector_model = torch.jit.trace(pose_landmark_detector_model, [torch.tensor(data[0]) for _, data in pose_landmark_detector_sample_inputs.items()])

# Compile model on a specific device
pose_landmark_detector_compile_job = hub.submit_compile_job(
    model=traced_pose_landmark_detector_model ,
    device=device,
    input_specs=pose_landmark_detector_model.get_input_spec(),
)

# Get target model to run on-device
pose_landmark_detector_target_model = pose_landmark_detector_compile_job.get_target_model()

```


Step 2: **Performance profiling on cloud-hosted device**

After compiling models from step 1. Models can be profiled model on-device using the
`target_model`. Note that this scripts runs the model on a device automatically
provisioned in the cloud.  Once the job is submitted, you can navigate to a
provided job URL to view a variety of on-device performance metrics.
```python
pose_detector_profile_job = hub.submit_profile_job(
    model=pose_detector_target_model,
    device=device,
)
pose_landmark_detector_profile_job = hub.submit_profile_job(
    model=pose_landmark_detector_target_model,
    device=device,
)

```

Step 3: **Verify on-device accuracy**

To verify the accuracy of the model on-device, you can run on-device inference
on sample input data on the same cloud hosted device.
```python
pose_detector_input_data = pose_detector_model.sample_inputs()
pose_detector_inference_job = hub.submit_inference_job(
    model=pose_detector_target_model,
    device=device,
    inputs=pose_detector_input_data,
)
pose_detector_inference_job.download_output_data()
pose_landmark_detector_input_data = pose_landmark_detector_model.sample_inputs()
pose_landmark_detector_inference_job = hub.submit_inference_job(
    model=pose_landmark_detector_target_model,
    device=device,
    inputs=pose_landmark_detector_input_data,
)
pose_landmark_detector_inference_job.download_output_data()

```
With the output of the model, you can compute like PSNR, relative errors or
spot check the output with expected output.

**Note**: This on-device profiling and inference requires access to Qualcomm®
AI Hub. [Sign up for access](https://myaccount.qualcomm.com/signup).




## 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 MediaPipe-Pose-Estimation's performance across various devices [here](https://aihub.qualcomm.com/models/mediapipe_pose).
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)


## License
* The license for the original implementation of MediaPipe-Pose-Estimation can be found [here](https://github.com/zmurez/MediaPipePyTorch/blob/master/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
* [BlazePose: On-device Real-time Body Pose tracking](https://arxiv.org/abs/2006.10204)
* [Source Model Implementation](https://github.com/zmurez/MediaPipePyTorch/)



## 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).