DDColor: Optimized for Qualcomm Devices

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

This is based on the implementation of DDColor 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 2.37, ONNX Runtime 1.23.0	Download
ONNX	w8a8	Universal	QAIRT 2.37, ONNX Runtime 1.23.0	Download
ONNX	w8a8_mixed_int16	Universal	QAIRT 2.37, ONNX Runtime 1.23.0	Download
QNN_DLC	float	Universal	QAIRT 2.42	Download
QNN_DLC	w8a8	Universal	QAIRT 2.42	Download
QNN_DLC	w8a8_mixed_int16	Universal	QAIRT 2.42	Download
TFLITE	float	Universal	QAIRT 2.42, TFLite 2.17.0	Download
TFLITE	w8a8	Universal	QAIRT 2.42, TFLite 2.17.0	Download

For more device-specific assets and performance metrics, visit DDColor 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 DDColor on GitHub for usage instructions.

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

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
DDColor	QNN_DLC	float	Snapdragon® X Elite	1135.963 ms	1 - 1 MB	NPU
DDColor	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	826.47 ms	0 - 1160 MB	NPU
DDColor	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	1983.976 ms	1 - 795 MB	NPU
DDColor	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	1100.967 ms	1 - 4 MB	NPU
DDColor	QNN_DLC	float	Qualcomm® SA8775P	1101.267 ms	1 - 644 MB	NPU
DDColor	QNN_DLC	float	Qualcomm® QCS9075	1106.511 ms	1 - 4 MB	NPU
DDColor	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	1191.042 ms	0 - 482 MB	NPU
DDColor	QNN_DLC	float	Qualcomm® SA7255P	1983.976 ms	1 - 795 MB	NPU
DDColor	QNN_DLC	float	Qualcomm® SA8295P	1222.255 ms	1 - 410 MB	NPU
DDColor	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	832.809 ms	1 - 754 MB	NPU
DDColor	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	682.665 ms	1 - 655 MB	NPU
DDColor	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	1206.113 ms	0 - 569 MB	NPU
DDColor	TFLITE	w8a8	Qualcomm® QCS6490	678.572 ms	106 - 235 MB	CPU
DDColor	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	3031.978 ms	1 - 483 MB	NPU
DDColor	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	1697.236 ms	0 - 4 MB	NPU
DDColor	TFLITE	w8a8	Qualcomm® SA8775P	1730.938 ms	0 - 496 MB	NPU
DDColor	TFLITE	w8a8	Qualcomm® QCS9075	1670.237 ms	0 - 62 MB	NPU
DDColor	TFLITE	w8a8	Qualcomm® QCM6690	1694.244 ms	9 - 409 MB	CPU
DDColor	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	2119.822 ms	0 - 551 MB	NPU
DDColor	TFLITE	w8a8	Qualcomm® SA7255P	3031.978 ms	1 - 483 MB	NPU
DDColor	TFLITE	w8a8	Qualcomm® SA8295P	2039.833 ms	0 - 348 MB	NPU
DDColor	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	936.183 ms	0 - 457 MB	NPU
DDColor	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	487.698 ms	4 - 374 MB	CPU
DDColor	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	936.444 ms	0 - 492 MB	NPU

License

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

References

• DDColor: Towards Photo-Realistic Image Colorization via Dual Decoders
• 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.