Yolo-v5: Optimized for Mobile Deployment

Real-time object detection optimized for mobile and edge

YoloV5 is a machine learning model that predicts bounding boxes and classes of objects in an image.

This model is an implementation of Yolo-v5 found here.

This repository provides scripts to run Yolo-v5 on Qualcomm® devices. More details on model performance across various devices, can be found here.

WARNING: The model assets are not readily available for download due to licensing restrictions.

Model Details

• Model Type: Model_use_case.object_detection
• Model Stats:
  • Model checkpoint: YoloV5-M
  • Input resolution: 640x640
  • Number of parameters: 21.2M
  • Model size (float): 81.1 MB
  • Model size (w8a16): 21.8 MB

Model	Precision	Device	Chipset	Target Runtime	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit	Target Model
Yolo-v5	float	QCS8275 (Proxy)	Qualcomm® QCS8275 (Proxy)	TFLITE	65.123 ms	0 - 73 MB	NPU	--
Yolo-v5	float	QCS8275 (Proxy)	Qualcomm® QCS8275 (Proxy)	QNN_DLC	63.605 ms	2 - 101 MB	NPU	--
Yolo-v5	float	QCS8450 (Proxy)	Qualcomm® QCS8450 (Proxy)	TFLITE	23.066 ms	0 - 95 MB	NPU	--
Yolo-v5	float	QCS8450 (Proxy)	Qualcomm® QCS8450 (Proxy)	QNN_DLC	26.414 ms	5 - 58 MB	NPU	--
Yolo-v5	float	QCS8550 (Proxy)	Qualcomm® QCS8550 (Proxy)	TFLITE	12.046 ms	0 - 15 MB	NPU	--
Yolo-v5	float	QCS8550 (Proxy)	Qualcomm® QCS8550 (Proxy)	QNN_DLC	11.145 ms	5 - 65 MB	NPU	--
Yolo-v5	float	QCS8550 (Proxy)	Qualcomm® QCS8550 (Proxy)	ONNX	13.326 ms	0 - 99 MB	NPU	--
Yolo-v5	float	QCS9075 (Proxy)	Qualcomm® QCS9075 (Proxy)	TFLITE	18.471 ms	0 - 73 MB	NPU	--
Yolo-v5	float	QCS9075 (Proxy)	Qualcomm® QCS9075 (Proxy)	QNN_DLC	17.325 ms	3 - 93 MB	NPU	--
Yolo-v5	float	Samsung Galaxy S24	Snapdragon® 8 Gen 3 Mobile	TFLITE	9.214 ms	0 - 109 MB	NPU	--
Yolo-v5	float	Samsung Galaxy S24	Snapdragon® 8 Gen 3 Mobile	QNN_DLC	8.394 ms	5 - 159 MB	NPU	--
Yolo-v5	float	Samsung Galaxy S24	Snapdragon® 8 Gen 3 Mobile	ONNX	9.799 ms	5 - 128 MB	NPU	--
Yolo-v5	float	Samsung Galaxy S25	Snapdragon® 8 Elite For Galaxy Mobile	TFLITE	7.161 ms	0 - 84 MB	NPU	--
Yolo-v5	float	Samsung Galaxy S25	Snapdragon® 8 Elite For Galaxy Mobile	QNN_DLC	6.307 ms	5 - 92 MB	NPU	--
Yolo-v5	float	Samsung Galaxy S25	Snapdragon® 8 Elite For Galaxy Mobile	ONNX	7.525 ms	0 - 89 MB	NPU	--
Yolo-v5	float	Snapdragon X Elite CRD	Snapdragon® X Elite	QNN_DLC	11.868 ms	108 - 108 MB	NPU	--
Yolo-v5	float	Snapdragon X Elite CRD	Snapdragon® X Elite	ONNX	13.574 ms	46 - 46 MB	NPU	--
Yolo-v5	w8a16	QCS8275 (Proxy)	Qualcomm® QCS8275 (Proxy)	QNN_DLC	21.145 ms	2 - 72 MB	NPU	--
Yolo-v5	w8a16	QCS8450 (Proxy)	Qualcomm® QCS8450 (Proxy)	QNN_DLC	13.496 ms	2 - 89 MB	NPU	--
Yolo-v5	w8a16	QCS8550 (Proxy)	Qualcomm® QCS8550 (Proxy)	QNN_DLC	8.886 ms	2 - 23 MB	NPU	--
Yolo-v5	w8a16	QCS9075 (Proxy)	Qualcomm® QCS9075 (Proxy)	QNN_DLC	9.068 ms	2 - 72 MB	NPU	--
Yolo-v5	w8a16	RB3 Gen 2 (Proxy)	Qualcomm® QCS6490 (Proxy)	QNN_DLC	55.179 ms	2 - 101 MB	NPU	--
Yolo-v5	w8a16	Samsung Galaxy S24	Snapdragon® 8 Gen 3 Mobile	QNN_DLC	5.894 ms	2 - 89 MB	NPU	--
Yolo-v5	w8a16	Samsung Galaxy S25	Snapdragon® 8 Elite For Galaxy Mobile	QNN_DLC	4.317 ms	2 - 77 MB	NPU	--
Yolo-v5	w8a16	Snapdragon X Elite CRD	Snapdragon® X Elite	QNN_DLC	9.554 ms	34 - 34 MB	NPU	--

Installation

Install the package via pip:

pip install "qai-hub-models[yolov5]"

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.yolov5.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.yolov5.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.yolov5.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.yolov5 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.yolov5.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.yolov5.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-v5's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

• The license for the original implementation of Yolo-v5 can be found here.
• The license for the compiled assets for on-device deployment can be found here

References

• Source Model Implementation

Community

• Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
• For questions or feedback please reach out to us.