SINet: Optimized for Qualcomm Devices

SINet is a machine learning model that is designed to segment people from close-up portrait images in real time.

This is based on the implementation of SINet 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 1.27.1	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a16	Universal	QAIRT 2.45	Download
QNN_DLC	w8a8	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

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

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

Input resolution: 224x224
Model checkpoint: SINet.pth
Model size (float): 415 KB
Model size (w8a8): 241 KB
Number of output classes: 2 (foreground / background)
Number of parameters: 91.9K

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
SINet	ONNX	float	Snapdragon® X2 Elite	0.762 ms	2 - 2 MB	NPU
SINet	ONNX	float	Snapdragon® X Elite	1.733 ms	2 - 2 MB	NPU
SINet	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.12 ms	0 - 47 MB	NPU
SINet	ONNX	float	Snapdragon® 8 Gen 1 Mobile	2.032 ms	1 - 57 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	1.84 ms	1 - 5 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.687 ms	0 - 4 MB	NPU
SINet	ONNX	float	Qualcomm® QCS8450	2.032 ms	1 - 57 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	1.909 ms	1 - 5 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	1.733 ms	2 - 2 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ Q-8750	0.841 ms	0 - 34 MB	NPU
SINet	ONNX	float	Snapdragon® 8 Elite Mobile	0.841 ms	0 - 34 MB	NPU
SINet	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	0.707 ms	0 - 37 MB	NPU
SINet	QNN_DLC	float	Snapdragon® X2 Elite	1.416 ms	1 - 1 MB	NPU
SINet	QNN_DLC	float	Snapdragon® X Elite	3.845 ms	1 - 1 MB	NPU
SINet	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.107 ms	0 - 46 MB	NPU
SINet	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	3.74 ms	0 - 55 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	3.062 ms	1 - 4 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	6.627 ms	1 - 34 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	3.431 ms	1 - 3 MB	NPU
SINet	QNN_DLC	float	Qualcomm® SA8775P	3.922 ms	1 - 36 MB	NPU
SINet	QNN_DLC	float	Qualcomm® SA8650P	3.922 ms	1 - 36 MB	NPU
SINet	QNN_DLC	float	Qualcomm® SA8255P	3.922 ms	1 - 36 MB	NPU
SINet	QNN_DLC	float	Qualcomm® QCS8450	3.74 ms	0 - 55 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	3.694 ms	1 - 4 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	3.845 ms	1 - 1 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	1.801 ms	1 - 33 MB	NPU
SINet	QNN_DLC	float	Qualcomm® SA7255P	6.627 ms	1 - 34 MB	NPU
SINet	QNN_DLC	float	Qualcomm® SA8295P	4.64 ms	1 - 36 MB	NPU
SINet	QNN_DLC	float	Snapdragon® 8 Elite Mobile	1.801 ms	1 - 33 MB	NPU
SINet	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.129 ms	1 - 36 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® X2 Elite	3.504 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® X Elite	7.156 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	4.079 ms	0 - 49 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® 8 Gen 1 Mobile	7.115 ms	0 - 55 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-8275	5.448 ms	0 - 3 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-8275	10.829 ms	0 - 38 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	6.582 ms	0 - 2 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® SA8775P	7.045 ms	0 - 41 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® SA8650P	7.045 ms	0 - 41 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® SA8255P	7.045 ms	0 - 41 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® QCS8450	7.115 ms	0 - 55 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	6.832 ms	2 - 5 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	7.156 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	3.438 ms	0 - 37 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® SA7255P	10.829 ms	0 - 38 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® SA8295P	8.141 ms	0 - 37 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	3.438 ms	0 - 37 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	3.209 ms	0 - 39 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® X2 Elite	2.346 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® X Elite	4.732 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	2.906 ms	0 - 47 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	4.631 ms	0 - 52 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-8275	4.036 ms	0 - 3 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-8275	7.736 ms	0 - 38 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	4.48 ms	0 - 2 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® SA8775P	4.706 ms	0 - 39 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® SA8650P	4.706 ms	0 - 39 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® SA8255P	4.706 ms	0 - 39 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® QCS8450	4.631 ms	0 - 52 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-9075	4.825 ms	0 - 2 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-X7181	4.732 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-8750	2.352 ms	0 - 36 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® SA7255P	7.736 ms	0 - 38 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® SA8295P	5.187 ms	0 - 35 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	2.352 ms	0 - 36 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	2.14 ms	0 - 38 MB	NPU
SINet	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.106 ms	0 - 45 MB	NPU
SINet	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	3.781 ms	0 - 54 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	3.077 ms	0 - 5 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	6.656 ms	1 - 35 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	3.405 ms	0 - 2 MB	NPU
SINet	TFLITE	float	Qualcomm® SA8775P	3.916 ms	0 - 38 MB	NPU
SINet	TFLITE	float	Qualcomm® SA8650P	3.916 ms	0 - 38 MB	NPU
SINet	TFLITE	float	Qualcomm® SA8255P	3.916 ms	0 - 38 MB	NPU
SINet	TFLITE	float	Qualcomm® QCS8450	3.781 ms	0 - 54 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	3.703 ms	1 - 4 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	1.806 ms	0 - 30 MB	NPU
SINet	TFLITE	float	Qualcomm® SA7255P	6.656 ms	1 - 35 MB	NPU
SINet	TFLITE	float	Qualcomm® SA8295P	4.661 ms	1 - 35 MB	NPU
SINet	TFLITE	float	Snapdragon® 8 Elite Mobile	1.806 ms	0 - 30 MB	NPU
SINet	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.137 ms	0 - 35 MB	NPU

License

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

References

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.

Downloads last month: -; Downloads are not tracked for this model. How to track

Paper for qualcomm/SINet

SINet: Extreme Lightweight Portrait Segmentation Networks with Spatial Squeeze Modules and Information Blocking Decoder

Paper • 1911.09099 • Published Nov 20, 2019