Hugging Face's logo

qualcomm
/
Posenet-Mobilenet

Keypoint Detection
PyTorch
TF Lite
ONNX
android
Model card Files Community
Posenet-Mobilenet / README.md
qaihm-bot's picture
qaihm-bot
Upload README.md with huggingface_hub
2c851bb verified
preview code
|
raw
history blame
10.8 kB
---
datasets:
- coco
library_name: pytorch
license: apache-2.0
pipeline_tag: keypoint-detection
tags:
- android
---
![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/posenet_mobilenet/web-assets/model_demo.png)
# Posenet-Mobilenet: Optimized for Mobile Deployment
## Perform accurate human pose estimation
Posenet performs pose estimation on human images.
This model is an implementation of Posenet-Mobilenet found [here]({source_repo}).
This repository provides scripts to run Posenet-Mobilenet on Qualcomm® devices.
More details on model performance across various devices, can be found
[here](https://aihub.qualcomm.com/models/posenet_mobilenet).
### Model Details
- **Model Type:** Pose estimation
- **Model Stats:**
- Model checkpoint: mobilenet_v1_101
- Input resolution: 513x257
- Number of parameters: 3.31M
- Model size: 12.7 MB
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
|---|---|---|---|---|---|---|---|---|
| Posenet-Mobilenet | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.375 ms | 0 - 32 MB | FP16 | NPU | [Posenet-Mobilenet.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.tflite) |
| Posenet-Mobilenet | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 1.442 ms | 0 - 12 MB | FP16 | NPU | [Posenet-Mobilenet.so](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.so) |
| Posenet-Mobilenet | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 1.894 ms | 0 - 7 MB | FP16 | NPU | [Posenet-Mobilenet.onnx](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.onnx) |
| Posenet-Mobilenet | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.102 ms | 0 - 40 MB | FP16 | NPU | [Posenet-Mobilenet.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.tflite) |
| Posenet-Mobilenet | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 1.157 ms | 2 - 18 MB | FP16 | NPU | [Posenet-Mobilenet.so](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.so) |
| Posenet-Mobilenet | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 1.493 ms | 0 - 45 MB | FP16 | NPU | [Posenet-Mobilenet.onnx](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.onnx) |
| Posenet-Mobilenet | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.396 ms | 0 - 1 MB | FP16 | NPU | [Posenet-Mobilenet.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.tflite) |
| Posenet-Mobilenet | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 1.387 ms | 2 - 3 MB | FP16 | NPU | Use Export Script |
| Posenet-Mobilenet | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 1.369 ms | 0 - 2 MB | FP16 | NPU | [Posenet-Mobilenet.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.tflite) |
| Posenet-Mobilenet | SA8255 (Proxy) | SA8255P Proxy | QNN | 1.396 ms | 2 - 3 MB | FP16 | NPU | Use Export Script |
| Posenet-Mobilenet | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 1.364 ms | 0 - 9 MB | FP16 | NPU | [Posenet-Mobilenet.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.tflite) |
| Posenet-Mobilenet | SA8775 (Proxy) | SA8775P Proxy | QNN | 1.39 ms | 2 - 3 MB | FP16 | NPU | Use Export Script |
| Posenet-Mobilenet | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 1.372 ms | 0 - 2 MB | FP16 | NPU | [Posenet-Mobilenet.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.tflite) |
| Posenet-Mobilenet | SA8650 (Proxy) | SA8650P Proxy | QNN | 1.399 ms | 2 - 3 MB | FP16 | NPU | Use Export Script |
| Posenet-Mobilenet | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 2.195 ms | 0 - 41 MB | FP16 | NPU | [Posenet-Mobilenet.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.tflite) |
| Posenet-Mobilenet | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 2.293 ms | 2 - 21 MB | FP16 | NPU | Use Export Script |
| Posenet-Mobilenet | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.963 ms | 0 - 22 MB | FP16 | NPU | [Posenet-Mobilenet.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.tflite) |
| Posenet-Mobilenet | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 1.077 ms | 2 - 15 MB | FP16 | NPU | Use Export Script |
| Posenet-Mobilenet | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 1.076 ms | 0 - 24 MB | FP16 | NPU | [Posenet-Mobilenet.onnx](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.onnx) |
| Posenet-Mobilenet | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 1.556 ms | 2 - 2 MB | FP16 | NPU | Use Export Script |
| Posenet-Mobilenet | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.147 ms | 7 - 7 MB | FP16 | NPU | [Posenet-Mobilenet.onnx](https://huggingface.co/qualcomm/Posenet-Mobilenet/blob/main/Posenet-Mobilenet.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.posenet_mobilenet.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.posenet_mobilenet.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.posenet_mobilenet.export
```
```
Profiling Results
------------------------------------------------------------
Posenet-Mobilenet
Device : Samsung Galaxy S23 (13)
Runtime : TFLITE
Estimated inference time (ms) : 1.4
Estimated peak memory usage (MB): [0, 32]
Total # Ops : 41
Compute Unit(s) : NPU (41 ops)
```
## How does this work?
This [export script](https://aihub.qualcomm.com/models/posenet_mobilenet/qai_hub_models/models/Posenet-Mobilenet/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.posenet_mobilenet import
# Load the model
# Device
device = hub.Device("Samsung Galaxy S23")
```
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
profile_job = hub.submit_profile_job(
model=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
input_data = torch_model.sample_inputs()
inference_job = hub.submit_inference_job(
model=target_model,
device=device,
inputs=input_data,
)
on_device_output = 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).
## Run demo on a cloud-hosted device
You can also run the demo on-device.
```bash
python -m qai_hub_models.models.posenet_mobilenet.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.posenet_mobilenet.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 Posenet-Mobilenet's performance across various devices [here](https://aihub.qualcomm.com/models/posenet_mobilenet).
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
## License
* The license for the original implementation of Posenet-Mobilenet can be found [here](https://github.com/rwightman/posenet-pytorch/blob/master/LICENSE.txt).
* 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
* [PersonLab: Person Pose Estimation and Instance Segmentation with a Bottom-Up, Part-Based, Geometric Embedding Model](https://arxiv.org/abs/1803.08225)
* [Source Model Implementation](https://github.com/rwightman/posenet-pytorch)
## 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).