Hugging Face's logo

qualcomm
/
Yolo-X

Object Detection
PyTorch
LiteRT
real_time
android
Model card Files
xet
Community

Yolo-X: Optimized for Mobile Deployment

Real-time object detection optimized for mobile and edge

YoloX is a machine learning model that predicts bounding boxes and classes of objects in an image.

This model is an implementation of Yolo-X found here.

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

Model Details

  • Model Type: Model_use_case.object_detection
  • Model Stats:
    • Model checkpoint: YoloX Small
    • Input resolution: 640x640
    • Number of parameters: 8.98M
    • Model size (float): 34.3 MB
    • Model size (w8a16): 9.53 MB
    • Model size (w8a8): 8.96 MB
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
Yolo-X float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 31.777 ms 0 - 38 MB NPU Yolo-X.tflite
Yolo-X float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 31.662 ms 4 - 70 MB NPU Yolo-X.dlc
Yolo-X float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 14.482 ms 0 - 52 MB NPU Yolo-X.tflite
Yolo-X float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 18.689 ms 5 - 45 MB NPU Yolo-X.dlc
Yolo-X float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 8.295 ms 0 - 10 MB NPU Yolo-X.tflite
Yolo-X float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 8.477 ms 5 - 32 MB NPU Yolo-X.dlc
Yolo-X float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) ONNX 16.643 ms 0 - 77 MB NPU Yolo-X.onnx.zip
Yolo-X float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 11.613 ms 0 - 38 MB NPU Yolo-X.tflite
Yolo-X float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 11.521 ms 1 - 69 MB NPU Yolo-X.dlc
Yolo-X float SA7255P ADP Qualcomm® SA7255P TFLITE 31.777 ms 0 - 38 MB NPU Yolo-X.tflite
Yolo-X float SA7255P ADP Qualcomm® SA7255P QNN_DLC 31.662 ms 4 - 70 MB NPU Yolo-X.dlc
Yolo-X float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 8.333 ms 0 - 9 MB NPU Yolo-X.tflite
Yolo-X float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 8.48 ms 5 - 40 MB NPU Yolo-X.dlc
Yolo-X float SA8295P ADP Qualcomm® SA8295P TFLITE 16.094 ms 0 - 47 MB NPU Yolo-X.tflite
Yolo-X float SA8295P ADP Qualcomm® SA8295P QNN_DLC 15.066 ms 5 - 50 MB NPU Yolo-X.dlc
Yolo-X float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 8.313 ms 0 - 11 MB NPU Yolo-X.tflite
Yolo-X float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 8.499 ms 5 - 41 MB NPU Yolo-X.dlc
Yolo-X float SA8775P ADP Qualcomm® SA8775P TFLITE 11.613 ms 0 - 38 MB NPU Yolo-X.tflite
Yolo-X float SA8775P ADP Qualcomm® SA8775P QNN_DLC 11.521 ms 1 - 69 MB NPU Yolo-X.dlc
Yolo-X float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 6.134 ms 0 - 48 MB NPU Yolo-X.tflite
Yolo-X float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 6.18 ms 5 - 106 MB NPU Yolo-X.dlc
Yolo-X float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 11.039 ms 5 - 117 MB NPU Yolo-X.onnx.zip
Yolo-X float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 4.73 ms 0 - 44 MB NPU Yolo-X.tflite
Yolo-X float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 4.965 ms 5 - 79 MB NPU Yolo-X.dlc
Yolo-X float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 9.322 ms 3 - 81 MB NPU Yolo-X.onnx.zip
Yolo-X float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile TFLITE 3.855 ms 0 - 43 MB NPU Yolo-X.tflite
Yolo-X float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 4.051 ms 0 - 76 MB NPU Yolo-X.dlc
Yolo-X float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 7.953 ms 5 - 92 MB NPU Yolo-X.onnx.zip
Yolo-X float Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 9.227 ms 32 - 32 MB NPU Yolo-X.dlc
Yolo-X float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 16.009 ms 14 - 14 MB NPU Yolo-X.onnx.zip
Yolo-X w8a16 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 15.568 ms 1 - 43 MB NPU Yolo-X.dlc
Yolo-X w8a16 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 9.28 ms 2 - 57 MB NPU Yolo-X.dlc
Yolo-X w8a16 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 7.964 ms 2 - 15 MB NPU Yolo-X.dlc
Yolo-X w8a16 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) ONNX 14.025 ms 0 - 40 MB NPU Yolo-X.onnx.zip
Yolo-X w8a16 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 37.864 ms 1 - 44 MB NPU Yolo-X.dlc
Yolo-X w8a16 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) QNN_DLC 27.454 ms 2 - 49 MB NPU Yolo-X.dlc
Yolo-X w8a16 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) ONNX 393.947 ms 106 - 122 MB CPU Yolo-X.onnx.zip
Yolo-X w8a16 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) ONNX 340.949 ms 106 - 112 MB CPU Yolo-X.onnx.zip
Yolo-X w8a16 SA7255P ADP Qualcomm® SA7255P QNN_DLC 15.568 ms 1 - 43 MB NPU Yolo-X.dlc
Yolo-X w8a16 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 7.956 ms 2 - 18 MB NPU Yolo-X.dlc
Yolo-X w8a16 SA8295P ADP Qualcomm® SA8295P QNN_DLC 10.037 ms 2 - 52 MB NPU Yolo-X.dlc
Yolo-X w8a16 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 7.952 ms 2 - 18 MB NPU Yolo-X.dlc
Yolo-X w8a16 SA8775P ADP Qualcomm® SA8775P QNN_DLC 37.864 ms 1 - 44 MB NPU Yolo-X.dlc
Yolo-X w8a16 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 5.18 ms 2 - 56 MB NPU Yolo-X.dlc
Yolo-X w8a16 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 9.801 ms 2 - 117 MB NPU Yolo-X.onnx.zip
Yolo-X w8a16 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 3.98 ms 2 - 53 MB NPU Yolo-X.dlc
Yolo-X w8a16 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 9.049 ms 2 - 85 MB NPU Yolo-X.onnx.zip
Yolo-X w8a16 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 3.417 ms 2 - 53 MB NPU Yolo-X.dlc
Yolo-X w8a16 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 19.77 ms 2 - 89 MB NPU Yolo-X.onnx.zip
Yolo-X w8a16 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 8.665 ms 8 - 8 MB NPU Yolo-X.dlc
Yolo-X w8a16 Snapdragon X Elite CRD Snapdragon® X Elite ONNX 14.539 ms 9 - 9 MB NPU Yolo-X.onnx.zip
Yolo-X w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 6.498 ms 0 - 32 MB NPU Yolo-X.tflite
Yolo-X w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 5.441 ms 1 - 36 MB NPU Yolo-X.dlc
Yolo-X w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 3.109 ms 0 - 50 MB NPU Yolo-X.tflite
Yolo-X w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 2.868 ms 1 - 51 MB NPU Yolo-X.dlc
Yolo-X w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 2.831 ms 0 - 38 MB NPU Yolo-X.tflite
Yolo-X w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 2.306 ms 1 - 13 MB NPU Yolo-X.dlc
Yolo-X w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) ONNX 8.812 ms 1 - 16 MB NPU Yolo-X.onnx.zip
Yolo-X w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 3.293 ms 0 - 32 MB NPU Yolo-X.tflite
Yolo-X w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 2.71 ms 1 - 36 MB NPU Yolo-X.dlc
Yolo-X w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) TFLITE 101.665 ms 8 - 41 MB GPU Yolo-X.tflite
Yolo-X w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) QNN_DLC 9.963 ms 0 - 40 MB NPU Yolo-X.dlc
Yolo-X w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) ONNX 90.927 ms 44 - 61 MB CPU Yolo-X.onnx.zip
Yolo-X w8a8 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) ONNX 82.63 ms 41 - 53 MB CPU Yolo-X.onnx.zip
Yolo-X w8a8 SA7255P ADP Qualcomm® SA7255P TFLITE 6.498 ms 0 - 32 MB NPU Yolo-X.tflite
Yolo-X w8a8 SA7255P ADP Qualcomm® SA7255P QNN_DLC 5.441 ms 1 - 36 MB NPU Yolo-X.dlc
Yolo-X w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 2.825 ms 0 - 37 MB NPU Yolo-X.tflite
Yolo-X w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 2.304 ms 1 - 16 MB NPU Yolo-X.dlc
Yolo-X w8a8 SA8295P ADP Qualcomm® SA8295P TFLITE 4.255 ms 0 - 38 MB NPU Yolo-X.tflite
Yolo-X w8a8 SA8295P ADP Qualcomm® SA8295P QNN_DLC 3.605 ms 1 - 43 MB NPU Yolo-X.dlc
Yolo-X w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 2.885 ms 0 - 37 MB NPU Yolo-X.tflite
Yolo-X w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 2.3 ms 1 - 12 MB NPU Yolo-X.dlc
Yolo-X w8a8 SA8775P ADP Qualcomm® SA8775P TFLITE 3.293 ms 0 - 32 MB NPU Yolo-X.tflite
Yolo-X w8a8 SA8775P ADP Qualcomm® SA8775P QNN_DLC 2.71 ms 1 - 36 MB NPU Yolo-X.dlc
Yolo-X w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 1.879 ms 0 - 53 MB NPU Yolo-X.tflite
Yolo-X w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 1.537 ms 1 - 49 MB NPU Yolo-X.dlc
Yolo-X w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 6.503 ms 1 - 100 MB NPU Yolo-X.onnx.zip
Yolo-X w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 1.516 ms 0 - 39 MB NPU Yolo-X.tflite
Yolo-X w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 1.121 ms 1 - 44 MB NPU Yolo-X.dlc
Yolo-X w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 6.043 ms 1 - 75 MB NPU Yolo-X.onnx.zip
Yolo-X w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile TFLITE 1.311 ms 0 - 34 MB NPU Yolo-X.tflite
Yolo-X w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 0.883 ms 1 - 43 MB NPU Yolo-X.dlc
Yolo-X w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 8.337 ms 1 - 76 MB NPU Yolo-X.onnx.zip
Yolo-X w8a8 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 2.586 ms 17 - 17 MB NPU Yolo-X.dlc
Yolo-X w8a8 Snapdragon X Elite CRD Snapdragon® X Elite ONNX 9.398 ms 8 - 8 MB NPU Yolo-X.onnx.zip
Yolo-X w8a8_mixed_int16 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 7.966 ms 1 - 38 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 3.617 ms 1 - 11 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 4.063 ms 1 - 37 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 SA7255P ADP Qualcomm® SA7255P QNN_DLC 7.966 ms 1 - 38 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 3.613 ms 1 - 9 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 3.634 ms 1 - 12 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 SA8775P ADP Qualcomm® SA8775P QNN_DLC 4.063 ms 1 - 37 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 2.414 ms 1 - 51 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 1.753 ms 1 - 45 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 1.376 ms 1 - 49 MB NPU Yolo-X.dlc
Yolo-X w8a8_mixed_int16 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 4.025 ms 12 - 12 MB NPU Yolo-X.dlc

Installation

Install the package via pip:

pip install "qai-hub-models[yolox]" git+https://github.com/Megvii-BaseDetection/YOLOX.git@6ddff48 --no-build-isolation --use-pep517

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.yolox.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.yolox.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.yolox.export

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.yolox import Model

# Load the model
torch_model = Model.from_pretrained()

# Device
device = hub.Device("Samsung Galaxy S25")

# 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.yolox.demo --eval-mode 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.yolox.demo -- --eval-mode 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 Yolo-X's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

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

References

Community

Downloads last month
733
Inference Providers NEW
Object Detection
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support