ESRGAN: Optimized for Qualcomm Devices

ESRGAN is a machine learning model that upscales an image with minimal loss in quality.

This is based on the implementation of ESRGAN 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 2.42, ONNX Runtime 1.24.3	Download
ONNX	w8a16	Universal	QAIRT 2.42, ONNX Runtime 1.24.3	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 ESRGAN 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 ESRGAN on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.super_resolution

Model Stats:

• Model checkpoint: ESRGAN_x4
• Input resolution: 128x128
• Number of parameters: 16.7M
• Model size (float): 63.9 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
ESRGAN	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	28.089 ms	7 - 353 MB	NPU
ESRGAN	ONNX	float	Snapdragon® 8 Elite Mobile	38.221 ms	0 - 332 MB	NPU
ESRGAN	ONNX	float	Snapdragon® X2 Elite	34.445 ms	37 - 37 MB	NPU
ESRGAN	ONNX	float	Snapdragon® X Elite	65.495 ms	37 - 37 MB	NPU
ESRGAN	ONNX	float	Snapdragon® X Elite	65.495 ms	37 - 37 MB	NPU
ESRGAN	ONNX	float	Snapdragon® 8 Gen 3 Mobile	49.588 ms	6 - 789 MB	NPU
ESRGAN	ONNX	float	Qualcomm® QCS8550 (Proxy)	66.58 ms	0 - 43 MB	NPU
ESRGAN	ONNX	float	Qualcomm® QCS9075	107.369 ms	8 - 11 MB	NPU
ESRGAN	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	38.221 ms	0 - 332 MB	NPU
ESRGAN	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	16.954 ms	3 - 1057 MB	NPU
ESRGAN	ONNX	w8a16	Snapdragon® 8 Elite Mobile	26.585 ms	3 - 905 MB	NPU
ESRGAN	ONNX	w8a16	Snapdragon® X2 Elite	22.017 ms	29 - 29 MB	NPU
ESRGAN	ONNX	w8a16	Snapdragon® X Elite	43.583 ms	26 - 26 MB	NPU
ESRGAN	ONNX	w8a16	Snapdragon® X Elite	43.583 ms	26 - 26 MB	NPU
ESRGAN	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	31.664 ms	3 - 1272 MB	NPU
ESRGAN	ONNX	w8a16	Qualcomm® QCS6490	15009.872 ms	201 - 206 MB	CPU
ESRGAN	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	41.934 ms	0 - 36 MB	NPU
ESRGAN	ONNX	w8a16	Qualcomm® QCS9075	45.338 ms	3 - 6 MB	NPU
ESRGAN	ONNX	w8a16	Qualcomm® QCM6690	7959.049 ms	185 - 206 MB	CPU
ESRGAN	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	26.585 ms	3 - 905 MB	NPU
ESRGAN	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	7739.408 ms	200 - 222 MB	CPU
ESRGAN	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	7739.408 ms	200 - 222 MB	CPU
ESRGAN	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	24.682 ms	0 - 330 MB	NPU
ESRGAN	QNN_DLC	float	Snapdragon® 8 Elite Mobile	37.48 ms	0 - 334 MB	NPU
ESRGAN	QNN_DLC	float	Snapdragon® X2 Elite	34.348 ms	0 - 0 MB	NPU
ESRGAN	QNN_DLC	float	Snapdragon® X Elite	64.943 ms	0 - 0 MB	NPU
ESRGAN	QNN_DLC	float	Snapdragon® X Elite	64.943 ms	0 - 0 MB	NPU
ESRGAN	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	48.637 ms	0 - 702 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	452.076 ms	0 - 353 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	62.637 ms	0 - 3 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® SA8775P	105.507 ms	0 - 353 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® SA8775P	105.507 ms	0 - 353 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® SA8775P	105.507 ms	0 - 353 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® QCS9075	106.465 ms	0 - 5 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	110.655 ms	0 - 709 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® SA7255P	452.076 ms	0 - 353 MB	NPU
ESRGAN	QNN_DLC	float	Qualcomm® SA8295P	111.344 ms	0 - 359 MB	NPU
ESRGAN	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	37.48 ms	0 - 334 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	16.274 ms	0 - 992 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	28.447 ms	0 - 843 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® X2 Elite	21.869 ms	0 - 0 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® X Elite	43.108 ms	0 - 0 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® X Elite	43.108 ms	0 - 0 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	31.364 ms	0 - 1035 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® QCS6490	240.627 ms	0 - 3 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® QCS8275 (Proxy)	132.979 ms	0 - 672 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® QCS8550 (Proxy)	42.811 ms	0 - 2 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® SA8775P	37.655 ms	0 - 672 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® SA8775P	37.655 ms	0 - 672 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® SA8775P	37.655 ms	0 - 672 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® QCS9075	44.3 ms	0 - 3 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® QCM6690	1115.421 ms	0 - 630 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® QCS8450 (Proxy)	82.137 ms	0 - 1113 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® SA7255P	132.979 ms	0 - 672 MB	NPU
ESRGAN	QNN_DLC	w8a16	Qualcomm® SA8295P	65.076 ms	0 - 717 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	28.447 ms	0 - 843 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	89.041 ms	0 - 696 MB	NPU
ESRGAN	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	89.041 ms	0 - 696 MB	NPU
ESRGAN	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	26.128 ms	3 - 369 MB	NPU
ESRGAN	TFLITE	float	Snapdragon® 8 Elite Mobile	37.699 ms	3 - 375 MB	NPU
ESRGAN	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	48.749 ms	0 - 742 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® QCS8275 (Proxy)	452.032 ms	3 - 397 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® QCS8550 (Proxy)	65.821 ms	3 - 6 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® SA8775P	105.589 ms	3 - 395 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® SA8775P	105.589 ms	3 - 395 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® SA8775P	105.589 ms	3 - 395 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® QCS9075	108.985 ms	3 - 46 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® QCS8450 (Proxy)	118.24 ms	3 - 744 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® SA7255P	452.032 ms	3 - 397 MB	NPU
ESRGAN	TFLITE	float	Qualcomm® SA8295P	111.392 ms	3 - 391 MB	NPU
ESRGAN	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	37.699 ms	3 - 375 MB	NPU

License

• The license for the original implementation of ESRGAN can be found here.

References

• ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks
• 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.