README.md

---
library_name: pytorch
license: other
tags:
- real_time
- android
pipeline_tag: object-detection

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/yolov6/web-assets/model_demo.png)

# Yolo-v6: Optimized for Mobile Deployment
## Real-time object detection optimized for mobile and edge


YoloV6 is a machine learning model that predicts bounding boxes and classes of objects in an image.

This model is an implementation of Yolo-v6 found [here](https://github.com/meituan/YOLOv6/).


This repository provides scripts to run Yolo-v6 on Qualcomm® devices.
More details on model performance across various devices, can be found
[here](https://aihub.qualcomm.com/models/yolov6).

**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: YoloV6-N
  - Input resolution: 640x640
  - Number of parameters: 4.68M
  - Model size (float): 17.9 MB
  - Model size (w8a8): 4.68 MB
  - Model size (w8a16): 5.03 MB

| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|---|---|---|---|---|---|---|---|---|
| Yolo-v6 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 22.578 ms | 0 - 31 MB | NPU | -- |
| Yolo-v6 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 14.693 ms | 2 - 70 MB | NPU | -- |
| Yolo-v6 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 12.22 ms | 0 - 42 MB | NPU | -- |
| Yolo-v6 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 6.754 ms | 5 - 41 MB | NPU | -- |
| Yolo-v6 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 11.377 ms | 0 - 40 MB | NPU | -- |
| Yolo-v6 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 4.652 ms | 5 - 54 MB | NPU | -- |
| Yolo-v6 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 8.01 ms | 1 - 55 MB | NPU | -- |
| Yolo-v6 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 13.063 ms | 0 - 30 MB | NPU | -- |
| Yolo-v6 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 6.298 ms | 1 - 68 MB | NPU | -- |
| Yolo-v6 | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 22.578 ms | 0 - 31 MB | NPU | -- |
| Yolo-v6 | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 14.693 ms | 2 - 70 MB | NPU | -- |
| Yolo-v6 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 11.338 ms | 0 - 39 MB | NPU | -- |
| Yolo-v6 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 4.642 ms | 0 - 36 MB | NPU | -- |
| Yolo-v6 | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 13.234 ms | 0 - 37 MB | NPU | -- |
| Yolo-v6 | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 6.898 ms | 4 - 36 MB | NPU | -- |
| Yolo-v6 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 11.398 ms | 0 - 39 MB | NPU | -- |
| Yolo-v6 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 4.637 ms | 5 - 50 MB | NPU | -- |
| Yolo-v6 | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 13.063 ms | 0 - 30 MB | NPU | -- |
| Yolo-v6 | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 6.298 ms | 1 - 68 MB | NPU | -- |
| Yolo-v6 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 7.822 ms | 30 - 71 MB | NPU | -- |
| Yolo-v6 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 3.371 ms | 5 - 111 MB | NPU | -- |
| Yolo-v6 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 6.289 ms | 5 - 94 MB | NPU | -- |
| Yolo-v6 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 6.755 ms | 0 - 33 MB | NPU | -- |
| Yolo-v6 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 2.61 ms | 5 - 71 MB | NPU | -- |
| Yolo-v6 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 4.99 ms | 2 - 73 MB | NPU | -- |
| Yolo-v6 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 5.488 ms | 0 - 32 MB | NPU | -- |
| Yolo-v6 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 2.229 ms | 5 - 76 MB | NPU | -- |
| Yolo-v6 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 3.363 ms | 3 - 77 MB | NPU | -- |
| Yolo-v6 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 4.879 ms | 37 - 37 MB | NPU | -- |
| Yolo-v6 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 7.387 ms | 6 - 6 MB | NPU | -- |
| Yolo-v6 | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 5.129 ms | 1 - 31 MB | NPU | -- |
| Yolo-v6 | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 2.718 ms | 2 - 40 MB | NPU | -- |
| Yolo-v6 | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 2.115 ms | 1 - 19 MB | NPU | -- |
| Yolo-v6 | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 2.698 ms | 1 - 31 MB | NPU | -- |
| Yolo-v6 | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 8.665 ms | 2 - 37 MB | NPU | -- |
| Yolo-v6 | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 5.129 ms | 1 - 31 MB | NPU | -- |
| Yolo-v6 | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 2.123 ms | 1 - 17 MB | NPU | -- |
| Yolo-v6 | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 3.312 ms | 2 - 37 MB | NPU | -- |
| Yolo-v6 | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 2.122 ms | 1 - 18 MB | NPU | -- |
| Yolo-v6 | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 2.698 ms | 1 - 31 MB | NPU | -- |
| Yolo-v6 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 1.423 ms | 2 - 37 MB | NPU | -- |
| Yolo-v6 | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 1.09 ms | 2 - 42 MB | NPU | -- |
| Yolo-v6 | w8a16 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | QNN_DLC | 3.418 ms | 2 - 41 MB | NPU | -- |
| Yolo-v6 | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.914 ms | 2 - 42 MB | NPU | -- |
| Yolo-v6 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 2.41 ms | 10 - 10 MB | NPU | -- |
| Yolo-v6 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 4.479 ms | 0 - 26 MB | NPU | -- |
| Yolo-v6 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 3.19 ms | 1 - 27 MB | NPU | -- |
| Yolo-v6 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 2.247 ms | 0 - 35 MB | NPU | -- |
| Yolo-v6 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 1.692 ms | 1 - 42 MB | NPU | -- |
| Yolo-v6 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 2.069 ms | 0 - 30 MB | NPU | -- |
| Yolo-v6 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.37 ms | 1 - 29 MB | NPU | -- |
| Yolo-v6 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 2.503 ms | 0 - 26 MB | NPU | -- |
| Yolo-v6 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 1.702 ms | 1 - 28 MB | NPU | -- |
| Yolo-v6 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | TFLITE | 4.322 ms | 0 - 32 MB | NPU | -- |
| Yolo-v6 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 5.149 ms | 1 - 33 MB | NPU | -- |
| Yolo-v6 | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | TFLITE | 35.402 ms | 3 - 11 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 4.479 ms | 0 - 26 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 3.19 ms | 1 - 27 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 2.07 ms | 0 - 5 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.382 ms | 1 - 28 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 2.973 ms | 0 - 31 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 2.317 ms | 1 - 33 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 2.071 ms | 0 - 30 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.381 ms | 1 - 30 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 2.503 ms | 0 - 26 MB | NPU | -- |
| Yolo-v6 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 1.702 ms | 1 - 28 MB | NPU | -- |
| Yolo-v6 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 1.366 ms | 0 - 44 MB | NPU | -- |
| Yolo-v6 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.907 ms | 1 - 43 MB | NPU | -- |
| Yolo-v6 | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 1.084 ms | 0 - 29 MB | NPU | -- |
| Yolo-v6 | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.679 ms | 1 - 33 MB | NPU | -- |
| Yolo-v6 | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | TFLITE | 2.289 ms | 0 - 34 MB | NPU | -- |
| Yolo-v6 | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | QNN_DLC | 1.697 ms | 1 - 35 MB | NPU | -- |
| Yolo-v6 | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 1.015 ms | 0 - 32 MB | NPU | -- |
| Yolo-v6 | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.566 ms | 1 - 37 MB | NPU | -- |
| Yolo-v6 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 1.576 ms | 13 - 13 MB | NPU | -- |




## Installation


Install the package via pip:
```bash
# NOTE: 3.10 <= PYTHON_VERSION < 3.14 is supported.
pip install "qai-hub-models[yolov6]"
```


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

Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.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://workbench.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.yolov6.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.yolov6.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.yolov6.export
```



## How does this work?

This [export script](https://aihub.qualcomm.com/models/yolov6/qai_hub_models/models/Yolo-v6/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.yolov6 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.
```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 Workbench. [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.yolov6.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.yolov6.demo -- --eval-mode 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 Yolo-v6's performance across various devices [here](https://aihub.qualcomm.com/models/yolov6).
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)


## License
* The license for the original implementation of Yolo-v6 can be found
  [here](https://github.com/meituan/YOLOv6/blob/47625514e7480706a46ff3c0cd0252907ac12f22/LICENSE).
* The license for the compiled assets for on-device deployment can be found [here](https://github.com/meituan/YOLOv6/blob/47625514e7480706a46ff3c0cd0252907ac12f22/LICENSE)



## References
* [YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications](https://arxiv.org/abs/2209.02976)
* [Source Model Implementation](https://github.com/meituan/YOLOv6/)



## 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:[email protected]).