--- library_name: pytorch license: apache-2.0 tags: - quantized - android pipeline_tag: keypoint-detection --- ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/posenet_mobilenet_quantized/web-assets/model_demo.png) # Posenet-Mobilenet-Quantized: Optimized for Mobile Deployment ## Quantized human pose estimator Posenet performs pose estimation on human images. This model is an implementation of Posenet-Mobilenet-Quantized found [here](https://github.com/rwightman/posenet-pytorch). This repository provides scripts to run Posenet-Mobilenet-Quantized on Qualcomm® devices. More details on model performance across various devices, can be found [here](https://aihub.qualcomm.com/models/posenet_mobilenet_quantized). ### Model Details - **Model Type:** Pose estimation - **Model Stats:** - Model checkpoint: mobilenet_v1_101 - Input resolution: 513x257 - Number of parameters: 3.31M - Model size: 3.47 MB | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model |---|---|---|---|---|---|---|---|---| | Posenet-Mobilenet-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 0.561 ms | 0 - 13 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.561 ms | 0 - 3 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.so](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.so) | | Posenet-Mobilenet-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.394 ms | 0 - 28 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.397 ms | 0 - 19 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.so](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.so) | | Posenet-Mobilenet-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.462 ms | 0 - 25 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.465 ms | 0 - 25 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | SA7255P ADP | SA7255P | TFLITE | 7.825 ms | 0 - 16 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | SA7255P ADP | SA7255P | QNN | 8.069 ms | 0 - 8 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 0.57 ms | 0 - 12 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.563 ms | 0 - 3 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | SA8295P ADP | SA8295P | TFLITE | 1.206 ms | 0 - 18 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | SA8295P ADP | SA8295P | QNN | 1.592 ms | 0 - 10 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 0.568 ms | 0 - 12 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.56 ms | 0 - 3 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | SA8775P ADP | SA8775P | TFLITE | 0.981 ms | 0 - 15 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | SA8775P ADP | SA8775P | QNN | 1.133 ms | 0 - 8 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 2.168 ms | 0 - 22 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 3.021 ms | 0 - 12 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 14.147 ms | 0 - 7 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 7.825 ms | 0 - 16 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 8.069 ms | 0 - 8 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 0.559 ms | 0 - 5 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.563 ms | 0 - 3 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 0.981 ms | 0 - 15 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 1.133 ms | 0 - 8 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 0.729 ms | 0 - 22 MB | INT8 | NPU | [Posenet-Mobilenet-Quantized.tflite](https://huggingface.co/qualcomm/Posenet-Mobilenet-Quantized/blob/main/Posenet-Mobilenet-Quantized.tflite) | | Posenet-Mobilenet-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 0.854 ms | 0 - 24 MB | INT8 | NPU | Use Export Script | | Posenet-Mobilenet-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.687 ms | 0 - 0 MB | INT8 | NPU | Use Export Script | ## Installation Install the package via pip: ```bash pip install "qai-hub-models[posenet-mobilenet-quantized]" ``` ## 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_quantized.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_quantized.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_quantized.export ``` ``` Profiling Results ------------------------------------------------------------ Posenet-Mobilenet-Quantized Device : Samsung Galaxy S23 (13) Runtime : TFLITE Estimated inference time (ms) : 0.6 Estimated peak memory usage (MB): [0, 13] Total # Ops : 48 Compute Unit(s) : NPU (48 ops) ``` ## Run demo on a cloud-hosted device You can also run the demo on-device. ```bash python -m qai_hub_models.models.posenet_mobilenet_quantized.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_quantized.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-Quantized's performance across various devices [here](https://aihub.qualcomm.com/models/posenet_mobilenet_quantized). Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/) ## License * The license for the original implementation of Posenet-Mobilenet-Quantized 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).