Hugging Face's logo

qualcomm
/
DDColor

Image-to-Image
PyTorch
LiteRT
android
Model card Files
xet
Community

DDColor: Optimized for Mobile Deployment

Colorize image from the black-and-white image

DDColor is a coloring algorithm that produces natural, vivid color results from incoming black and white images.

This model is an implementation of DDColor found here.

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

Model Details

  • Model Type: Model_use_case.image_editing
  • Model Stats:
    • Model checkpoint: ddcolor_paper_tiny.pth
    • Input resolution: 224x224
    • Number of parameters: 56.3M
    • Model size (float): 215 MB
    • Model size (w8a8): 54.8 MB
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
DDColor float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 249.068 ms 0 - 332 MB NPU DDColor.tflite
DDColor float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 1989.568 ms 0 - 720 MB NPU DDColor.dlc
DDColor float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 169.0 ms 1 - 262 MB NPU DDColor.tflite
DDColor float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 1235.43 ms 1 - 253 MB NPU DDColor.dlc
DDColor float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 158.008 ms 0 - 34 MB NPU DDColor.tflite
DDColor float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 1124.836 ms 0 - 46 MB NPU DDColor.dlc
DDColor float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) ONNX 1127.711 ms 0 - 167 MB NPU DDColor.onnx.zip
DDColor float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 161.863 ms 1 - 347 MB NPU DDColor.tflite
DDColor float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 1112.75 ms 1 - 661 MB NPU DDColor.dlc
DDColor float SA7255P ADP Qualcomm® SA7255P TFLITE 249.068 ms 0 - 332 MB NPU DDColor.tflite
DDColor float SA7255P ADP Qualcomm® SA7255P QNN_DLC 1989.568 ms 0 - 720 MB NPU DDColor.dlc
DDColor float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 159.856 ms 0 - 34 MB NPU DDColor.tflite
DDColor float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 1127.688 ms 0 - 45 MB NPU DDColor.dlc
DDColor float SA8295P ADP Qualcomm® SA8295P TFLITE 176.33 ms 0 - 239 MB NPU DDColor.tflite
DDColor float SA8295P ADP Qualcomm® SA8295P QNN_DLC 1229.554 ms 0 - 394 MB NPU DDColor.dlc
DDColor float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 159.691 ms 0 - 35 MB NPU DDColor.tflite
DDColor float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 1120.366 ms 0 - 48 MB NPU DDColor.dlc
DDColor float SA8775P ADP Qualcomm® SA8775P TFLITE 161.863 ms 1 - 347 MB NPU DDColor.tflite
DDColor float SA8775P ADP Qualcomm® SA8775P QNN_DLC 1112.75 ms 1 - 661 MB NPU DDColor.dlc
DDColor float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 112.938 ms 1 - 356 MB NPU DDColor.tflite
DDColor float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 839.748 ms 0 - 837 MB NPU DDColor.dlc
DDColor float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 846.043 ms 1 - 952 MB NPU DDColor.onnx.zip
DDColor float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 95.693 ms 1 - 309 MB NPU DDColor.tflite
DDColor float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 839.94 ms 1 - 433 MB NPU DDColor.dlc
DDColor float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 859.57 ms 1 - 565 MB NPU DDColor.onnx.zip
DDColor float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile TFLITE 76.207 ms 0 - 325 MB NPU DDColor.tflite
DDColor float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 700.968 ms 0 - 608 MB NPU DDColor.dlc
DDColor float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 690.838 ms 1 - 673 MB NPU DDColor.onnx.zip
DDColor float Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 1179.236 ms 81 - 81 MB NPU DDColor.dlc
DDColor float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 1157.963 ms 113 - 113 MB NPU DDColor.onnx.zip
DDColor w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 3336.441 ms 0 - 362 MB NPU DDColor.tflite
DDColor w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 5187.205 ms 0 - 392 MB NPU DDColor.dlc
DDColor w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 1772.082 ms 0 - 395 MB NPU DDColor.tflite
DDColor w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 2974.197 ms 1 - 290 MB NPU DDColor.dlc
DDColor w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 1760.119 ms 0 - 53 MB NPU DDColor.tflite
DDColor w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 2717.716 ms 3 - 65 MB NPU DDColor.dlc
DDColor w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 1784.176 ms 0 - 362 MB NPU DDColor.tflite
DDColor w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 2832.254 ms 0 - 397 MB NPU DDColor.dlc
DDColor w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) TFLITE 614.783 ms 86 - 116 MB CPU DDColor.tflite
DDColor w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) ONNX 647.012 ms 328 - 355 MB CPU DDColor.onnx.zip
DDColor w8a8 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) TFLITE 817.267 ms 63 - 104 MB CPU DDColor.tflite
DDColor w8a8 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) ONNX 526.175 ms 325 - 349 MB CPU DDColor.onnx.zip
DDColor w8a8 SA7255P ADP Qualcomm® SA7255P TFLITE 3336.441 ms 0 - 362 MB NPU DDColor.tflite
DDColor w8a8 SA7255P ADP Qualcomm® SA7255P QNN_DLC 5187.205 ms 0 - 392 MB NPU DDColor.dlc
DDColor w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 1761.455 ms 0 - 47 MB NPU DDColor.tflite
DDColor w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 2717.403 ms 0 - 64 MB NPU DDColor.dlc
DDColor w8a8 SA8295P ADP Qualcomm® SA8295P TFLITE 2019.923 ms 0 - 370 MB NPU DDColor.tflite
DDColor w8a8 SA8295P ADP Qualcomm® SA8295P QNN_DLC 3156.066 ms 0 - 286 MB NPU DDColor.dlc
DDColor w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 1762.868 ms 0 - 24 MB NPU DDColor.tflite
DDColor w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 2716.067 ms 0 - 66 MB NPU DDColor.dlc
DDColor w8a8 SA8775P ADP Qualcomm® SA8775P TFLITE 1784.176 ms 0 - 362 MB NPU DDColor.tflite
DDColor w8a8 SA8775P ADP Qualcomm® SA8775P QNN_DLC 2832.254 ms 0 - 397 MB NPU DDColor.dlc
DDColor w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 1337.685 ms 0 - 381 MB NPU DDColor.tflite
DDColor w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 2046.081 ms 0 - 440 MB NPU DDColor.dlc
DDColor w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 1048.701 ms 0 - 106 MB NPU DDColor.tflite
DDColor w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 1786.258 ms 0 - 285 MB NPU DDColor.dlc
DDColor w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 1947.273 ms 47 - 256 MB NPU DDColor.onnx.zip
DDColor w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile TFLITE 925.38 ms 0 - 111 MB NPU DDColor.tflite
DDColor w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 1500.827 ms 0 - 423 MB NPU DDColor.dlc
DDColor w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 1608.329 ms 38 - 290 MB NPU DDColor.onnx.zip
DDColor w8a8 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 2827.194 ms 175 - 175 MB NPU DDColor.dlc
DDColor w8a8 Snapdragon X Elite CRD Snapdragon® X Elite ONNX 3266.117 ms 69 - 69 MB NPU DDColor.onnx.zip

Installation

Install the package via pip:

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

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.ddcolor.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.ddcolor.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.ddcolor.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.ddcolor 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.ddcolor.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.ddcolor.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 DDColor's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

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

References

Community

Downloads last month
46
Inference Providers NEW
Image-to-Image
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support