Hugging Face's logo

qualcomm
/
LiteHRNet

Keypoint Detection
PyTorch
LiteRT
ONNX
android
Model card Files Community

LiteHRNet: Optimized for Mobile Deployment

Human pose estimation

LiteHRNet is a machine learning model that detects human pose and returns a location and confidence for each of 17 joints.

This model is an implementation of LiteHRNet found here.

This repository provides scripts to run LiteHRNet on Qualcomm® devices. More details on model performance across various devices, can be found here.

Model Details

  • Model Type: Model_use_case.pose_estimation
  • Model Stats:
    • Input resolution: 256x192
    • Number of parameters: 1.11M
    • Model size: 4.56 MB
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
LiteHRNet float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 9.016 ms 0 - 51 MB NPU LiteHRNet.tflite
LiteHRNet float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 5.4 ms 0 - 61 MB NPU LiteHRNet.tflite
LiteHRNet float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 4.599 ms 0 - 14 MB NPU LiteHRNet.tflite
LiteHRNet float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 5.629 ms 0 - 52 MB NPU LiteHRNet.tflite
LiteHRNet float SA7255P ADP Qualcomm® SA7255P TFLITE 9.016 ms 0 - 51 MB NPU LiteHRNet.tflite
LiteHRNet float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 4.604 ms 0 - 15 MB NPU LiteHRNet.tflite
LiteHRNet float SA8295P ADP Qualcomm® SA8295P TFLITE 6.353 ms 0 - 54 MB NPU LiteHRNet.tflite
LiteHRNet float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 4.606 ms 0 - 15 MB NPU LiteHRNet.tflite
LiteHRNet float SA8775P ADP Qualcomm® SA8775P TFLITE 5.629 ms 0 - 52 MB NPU LiteHRNet.tflite
LiteHRNet float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile TFLITE 4.602 ms 0 - 15 MB NPU LiteHRNet.tflite
LiteHRNet float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile ONNX 7.375 ms 0 - 25 MB NPU LiteHRNet.onnx
LiteHRNet float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 2.847 ms 0 - 59 MB NPU LiteHRNet.tflite
LiteHRNet float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 4.592 ms 1 - 56 MB NPU LiteHRNet.onnx
LiteHRNet float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile TFLITE 2.869 ms 0 - 52 MB NPU LiteHRNet.tflite
LiteHRNet float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile ONNX 4.149 ms 1 - 52 MB NPU LiteHRNet.onnx
LiteHRNet float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 7.993 ms 4 - 4 MB NPU LiteHRNet.onnx

Installation

Install the package via pip:

pip install "qai-hub-models[litehrnet]" torch==2.4.1 -f https://download.openmmlab.com/mmcv/dist/cpu/torch2.4/index.html -f https://qaihub-public-python-wheels.s3.us-west-2.amazonaws.com/index.html

Configure Qualcomm® AI Hub to run this model on a cloud-hosted device

Sign-in to Qualcomm® AI Hub 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.

qai-hub configure --api_token API_TOKEN

Navigate to 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.

python -m qai_hub_models.models.litehrnet.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.litehrnet.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.
python -m qai_hub_models.models.litehrnet.export

Profiling Results
------------------------------------------------------------
LiteHRNet
Device                          : cs_8275 (ANDROID 14)                  
Runtime                         : TFLITE                                
Estimated inference time (ms)   : 9.0                                   
Estimated peak memory usage (MB): [0, 51]                               
Total # Ops                     : 1113                                  
Compute Unit(s)                 : npu (1111 ops) gpu (0 ops) cpu (2 ops)

How does this work?

This export script leverages Qualcomm® AI Hub 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.

import torch

import qai_hub as hub
from qai_hub_models.models.litehrnet import Model

# Load the model
torch_model = Model.from_pretrained()

# Device
device = hub.Device("Samsung Galaxy S24")

# Trace model
input_shape = torch_model.get_input_spec()
sample_inputs = torch_model.sample_inputs()

pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])

# Compile model on a specific device
compile_job = hub.submit_compile_job(
    model=pt_model,
    device=device,
    input_specs=torch_model.get_input_spec(),
)

# Get target model to run on-device
target_model = 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.

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.

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.

Run demo on a cloud-hosted device

You can also run the demo on-device.

python -m qai_hub_models.models.litehrnet.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.litehrnet.demo -- --on-device

Deploying compiled model to Android

The models can be deployed using multiple runtimes:

  • TensorFlow Lite (.tflite export): This tutorial provides a guide to deploy the .tflite model in an Android application.

  • QNN (.so export ): This sample app provides instructions on how to use the .so shared library in an Android application.

View on Qualcomm® AI Hub

Get more details on LiteHRNet's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

  • The license for the original implementation of LiteHRNet can be found here.
  • The license for the compiled assets for on-device deployment can be found here

References

Community

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Keypoint Detection
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support