README.md

---
library_name: pytorch
license: other
tags:
- generative_ai
- android
pipeline_tag: unconditional-image-generation

---

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

# ControlNet: Optimized for Mobile Deployment
## Generating visual arts from text prompt and input guiding image


On-device, high-resolution image synthesis from text and image prompts. ControlNet guides Stable-diffusion with provided input image to generate accurate images from given input prompt.

This model is an implementation of ControlNet found [here](https://github.com/lllyasviel/ControlNet).


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


### Model Details

- **Model Type:** Model_use_case.image_generation
- **Model Stats:**
  - Input: Text prompt and input image as a reference
  - Conditioning Input: Canny-Edge
  - Text Encoder Number of parameters: 340M
  - UNet Number of parameters: 865M
  - VAE Decoder Number of parameters: 83M
  - ControlNet Number of parameters: 361M
  - Model size: 1.4GB

| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|---|---|---|---|---|---|---|---|---|
| TextEncoder_Quantized | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 10.874 ms | 0 - 3 MB | NPU | Use Export Script |
| TextEncoder_Quantized | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 7.918 ms | 0 - 18 MB | NPU | Use Export Script |
| TextEncoder_Quantized | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 10.875 ms | 0 - 3 MB | NPU | Use Export Script |
| UNet_Quantized | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 258.151 ms | 13 - 15 MB | NPU | Use Export Script |
| UNet_Quantized | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 197.629 ms | 13 - 31 MB | NPU | Use Export Script |
| UNet_Quantized | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 256.936 ms | 13 - 16 MB | NPU | Use Export Script |
| VAEDecoder_Quantized | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 397.625 ms | 0 - 2 MB | NPU | Use Export Script |
| VAEDecoder_Quantized | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 300.627 ms | 0 - 21 MB | NPU | Use Export Script |
| VAEDecoder_Quantized | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 395.006 ms | 0 - 3 MB | NPU | Use Export Script |
| ControlNet_Quantized | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 104.668 ms | 2 - 9 MB | NPU | Use Export Script |
| ControlNet_Quantized | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 77.289 ms | 2 - 23 MB | NPU | Use Export Script |
| ControlNet_Quantized | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 103.817 ms | 2 - 5 MB | NPU | Use Export Script |




## Installation


Install the package via pip:
```bash
pip install "qai-hub-models[controlnet]"
```


## 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 on-device

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.controlnet.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.controlnet.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.controlnet.export
```
```
Profiling Results
------------------------------------------------------------
TextEncoder_Quantized
Device                          : cs_8_gen_2 (ANDROID 13)              
Runtime                         : QNN                                  
Estimated inference time (ms)   : 10.9                                 
Estimated peak memory usage (MB): [0, 3]                               
Total # Ops                     : 569                                  
Compute Unit(s)                 : npu (569 ops) gpu (0 ops) cpu (0 ops)

------------------------------------------------------------
UNet_Quantized
Device                          : cs_8_gen_2 (ANDROID 13)               
Runtime                         : QNN                                   
Estimated inference time (ms)   : 258.2                                 
Estimated peak memory usage (MB): [13, 15]                              
Total # Ops                     : 5433                                  
Compute Unit(s)                 : npu (5433 ops) gpu (0 ops) cpu (0 ops)

------------------------------------------------------------
VAEDecoder_Quantized
Device                          : cs_8_gen_2 (ANDROID 13)              
Runtime                         : QNN                                  
Estimated inference time (ms)   : 397.6                                
Estimated peak memory usage (MB): [0, 2]                               
Total # Ops                     : 408                                  
Compute Unit(s)                 : npu (408 ops) gpu (0 ops) cpu (0 ops)

------------------------------------------------------------
ControlNet_Quantized
Device                          : cs_8_gen_2 (ANDROID 13)               
Runtime                         : QNN                                   
Estimated inference time (ms)   : 104.7                                 
Estimated peak memory usage (MB): [2, 9]                                
Total # Ops                     : 2405                                  
Compute Unit(s)                 : npu (2405 ops) gpu (0 ops) cpu (0 ops)
```


## How does this work?

This [export script](https://aihub.qualcomm.com/models/controlnet/qai_hub_models/models/ControlNet/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: **Upload compiled model**

Upload compiled models from `qai_hub_models.models.controlnet` on hub.
```python
import torch

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

# Load the model
model = Model.from_precompiled()

```


Step 2: **Performance profiling on cloud-hosted device**

After uploading compiled 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

# Device
device = hub.Device("Samsung Galaxy S23")

```

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

```
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](https://myaccount.qualcomm.com/signup).




## 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` / `.bin` 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 or `.bin` context binary in an Android application.


## View on Qualcomm® AI Hub
Get more details on ControlNet's performance across various devices [here](https://aihub.qualcomm.com/models/controlnet).
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)


## License
* The license for the original implementation of ControlNet can be found
  [here](https://github.com/lllyasviel/ControlNet/blob/main/LICENSE).
* 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
* [Adding Conditional Control to Text-to-Image Diffusion Models](https://arxiv.org/abs/2302.05543)
* [Source Model Implementation](https://github.com/lllyasviel/ControlNet)



## 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).


## Usage and Limitations

This model may not be used for or in connection with any of the following applications:

- Accessing essential private and public services and benefits;
- Administration of justice and democratic processes;
- Assessing or recognizing the emotional state of a person;
- Biometric and biometrics-based systems, including categorization of persons based on sensitive characteristics;
- Education and vocational training;
- Employment and workers management;
- Exploitation of the vulnerabilities of persons resulting in harmful behavior;
- General purpose social scoring;
- Law enforcement;
- Management and operation of critical infrastructure;
- Migration, asylum and border control management;
- Predictive policing;
- Real-time remote biometric identification in public spaces;
- Recommender systems of social media platforms;
- Scraping of facial images (from the internet or otherwise); and/or
- Subliminal manipulation