Hugging Face's logo

qualcomm
/
NAFNet-DeNoise

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

NAFNet-DeNoise: Optimized for Qualcomm Devices

NAFNET is designed for lightweight real-time denoising of images.

This is based on the implementation of NAFNet-DeNoise found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up to run these models on a hosted Qualcomm® device.

Getting Started

There are two ways to deploy this model on your device:

Option 1: Download Pre-Exported Models

Below are pre-exported model assets ready for deployment.

Runtime Precision Chipset SDK Versions Download
ONNX float Universal QAIRT 2.45, ONNX Runtime 1.25.0 Download
ONNX w8a16 Universal QAIRT 2.45, ONNX Runtime 1.25.0 Download
QNN_DLC float Universal QAIRT 2.45 Download
QNN_DLC w8a16 Universal QAIRT 2.45 Download
TFLITE float Universal QAIRT 2.45 Download

For more device-specific assets and performance metrics, visit NAFNet-DeNoise on Qualcomm® AI Hub.

Option 2: Export with Custom Configurations

Use the Qualcomm® AI Hub Models Python library to compile and export the model with your own:

  • Custom weights (e.g., fine-tuned checkpoints)
  • Custom input shapes
  • Target device and runtime configurations

This option is ideal if you need to customize the model beyond the default configuration provided here.

See our repository for NAFNet-DeNoise on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.image_editing

Model Stats:

  • Model checkpoint: NAFNet-SIDD-width64
  • Input resolution: 256x256
  • Number of parameters: 115.98M
  • Model size (float): 463.93 MB

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
NAFNet-DeNoise ONNX float Snapdragon® X2 Elite 14.879 ms 180 - 180 MB NPU
NAFNet-DeNoise ONNX float Snapdragon® X Elite 33.765 ms 227 - 227 MB NPU
NAFNet-DeNoise ONNX float Snapdragon® 8 Gen 3 Mobile 27.393 ms 48 - 884 MB NPU
NAFNet-DeNoise ONNX float Snapdragon® 8 Gen 1 Mobile 61.732 ms 3 - 710 MB NPU
NAFNet-DeNoise ONNX float Qualcomm® QCS8550 (Proxy) 35.093 ms 0 - 260 MB NPU
NAFNet-DeNoise ONNX float Qualcomm® QCS8450 61.732 ms 3 - 710 MB NPU
NAFNet-DeNoise ONNX float Snapdragon® 8 Elite Mobile 20.1 ms 2 - 498 MB NPU
NAFNet-DeNoise ONNX float Snapdragon® 8 Elite Gen 5 Mobile 14.387 ms 2 - 533 MB NPU
NAFNet-DeNoise ONNX float Qualcomm® QCS9075 46.176 ms 2 - 47 MB NPU
NAFNet-DeNoise ONNX float Qualcomm® QCS8750 20.1 ms 2 - 498 MB NPU
NAFNet-DeNoise ONNX float Qualcomm® QCS7181 33.765 ms 227 - 227 MB NPU
NAFNet-DeNoise ONNX w8a16 Snapdragon® X2 Elite 14.404 ms 180 - 180 MB NPU
NAFNet-DeNoise ONNX w8a16 Snapdragon® X Elite 33.211 ms 150 - 150 MB NPU
NAFNet-DeNoise ONNX w8a16 Snapdragon® 8 Gen 3 Mobile 24.219 ms 1 - 687 MB NPU
NAFNet-DeNoise ONNX w8a16 Qualcomm® QCS8550 (Proxy) 33.004 ms 0 - 133 MB NPU
NAFNet-DeNoise ONNX w8a16 Qualcomm® QCS9075 45.675 ms 1 - 46 MB NPU
NAFNet-DeNoise ONNX w8a16 Snapdragon® 8 Elite Gen 5 Mobile 12.924 ms 1 - 681 MB NPU
NAFNet-DeNoise ONNX w8a16 Snapdragon® 8 Elite Mobile 19.178 ms 1 - 565 MB NPU
NAFNet-DeNoise ONNX w8a16 Snapdragon® 7 Gen 4 Mobile 38.656 ms 1 - 723 MB NPU
NAFNet-DeNoise ONNX w8a16 Qualcomm® QCM6690 296.283 ms 1 - 913 MB NPU
NAFNet-DeNoise ONNX w8a16 Qualcomm® QCS7790 38.656 ms 1 - 723 MB NPU
NAFNet-DeNoise ONNX w8a16 Qualcomm® QCS8750 19.178 ms 1 - 565 MB NPU
NAFNet-DeNoise ONNX w8a16 Qualcomm® QCS7181 33.211 ms 150 - 150 MB NPU
NAFNet-DeNoise QNN_DLC float Snapdragon® X2 Elite 15.724 ms 1 - 1 MB NPU
NAFNet-DeNoise QNN_DLC float Snapdragon® X Elite 35.797 ms 1 - 1 MB NPU
NAFNet-DeNoise QNN_DLC float Snapdragon® 8 Gen 3 Mobile 28.329 ms 0 - 889 MB NPU
NAFNet-DeNoise QNN_DLC float Snapdragon® 8 Gen 1 Mobile 60.451 ms 1 - 728 MB NPU
NAFNet-DeNoise QNN_DLC float Qualcomm® QCS8275 136.049 ms 1 - 546 MB NPU
NAFNet-DeNoise QNN_DLC float Qualcomm® QCS8550 (Proxy) 36.624 ms 1 - 369 MB NPU
NAFNet-DeNoise QNN_DLC float Qualcomm® QCS8450 60.451 ms 1 - 728 MB NPU
NAFNet-DeNoise QNN_DLC float Snapdragon® 8 Elite Mobile 20.596 ms 0 - 542 MB NPU
NAFNet-DeNoise QNN_DLC float Qualcomm® SA8295P 47.967 ms 1 - 399 MB NPU
NAFNet-DeNoise QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 14.667 ms 0 - 621 MB NPU
NAFNet-DeNoise QNN_DLC float Qualcomm® SA7255P 136.049 ms 1 - 546 MB NPU
NAFNet-DeNoise QNN_DLC float Qualcomm® QCS9075 46.521 ms 1 - 4 MB NPU
NAFNet-DeNoise QNN_DLC float Qualcomm® QCS8750 20.596 ms 0 - 542 MB NPU
NAFNet-DeNoise QNN_DLC float Qualcomm® QCS7181 35.797 ms 1 - 1 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Snapdragon® X2 Elite 15.099 ms 0 - 0 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Snapdragon® X Elite 33.008 ms 0 - 0 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Snapdragon® 8 Gen 3 Mobile 22.683 ms 0 - 687 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Qualcomm® QCS8275 65.109 ms 1 - 576 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Qualcomm® QCS8550 (Proxy) 31.446 ms 1 - 3 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Qualcomm® QCS9075 45.867 ms 0 - 3 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Snapdragon® 8 Elite Gen 5 Mobile 13.169 ms 0 - 706 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Snapdragon® 8 Elite Mobile 21.08 ms 0 - 586 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Snapdragon® 7 Gen 4 Mobile 39.247 ms 0 - 815 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Qualcomm® QCM6690 360.0 ms 0 - 953 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Qualcomm® SA7255P 65.109 ms 1 - 576 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Qualcomm® QCS7790 39.247 ms 0 - 815 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Qualcomm® QCS8750 21.08 ms 0 - 586 MB NPU
NAFNet-DeNoise QNN_DLC w8a16 Qualcomm® QCS7181 33.008 ms 0 - 0 MB NPU
NAFNet-DeNoise TFLITE float Snapdragon® 8 Gen 3 Mobile 28.269 ms 1 - 1120 MB NPU
NAFNet-DeNoise TFLITE float Snapdragon® 8 Gen 1 Mobile 64.655 ms 1 - 962 MB NPU
NAFNet-DeNoise TFLITE float Qualcomm® QCS8275 136.629 ms 1 - 765 MB NPU
NAFNet-DeNoise TFLITE float Qualcomm® QCS8550 (Proxy) 36.903 ms 1 - 4 MB NPU
NAFNet-DeNoise TFLITE float Qualcomm® SA8775P 292.421 ms 1 - 43 MB GPU
NAFNet-DeNoise TFLITE float Qualcomm® SA8650P 292.421 ms 1 - 43 MB GPU
NAFNet-DeNoise TFLITE float Qualcomm® SA8255P 292.421 ms 1 - 43 MB GPU
NAFNet-DeNoise TFLITE float Qualcomm® QCS8450 64.655 ms 1 - 962 MB NPU
NAFNet-DeNoise TFLITE float Snapdragon® 8 Elite Mobile 20.444 ms 1 - 760 MB NPU
NAFNet-DeNoise TFLITE float Qualcomm® SA8295P 50.802 ms 1 - 629 MB NPU
NAFNet-DeNoise TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 14.709 ms 1 - 850 MB NPU
NAFNet-DeNoise TFLITE float Qualcomm® SA7255P 136.629 ms 1 - 765 MB NPU
NAFNet-DeNoise TFLITE float Qualcomm® QCS9075 46.095 ms 1 - 232 MB NPU
NAFNet-DeNoise TFLITE float Qualcomm® QCS8750 20.444 ms 1 - 760 MB NPU

License

  • The license for the original implementation of NAFNet-DeNoise can be found here.

References

Community

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Image-to-Image
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Paper for qualcomm/NAFNet-DeNoise