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 2.50, ONNX Runtime 1.27.1	Download
QNN_DLC	float	Universal	QAIRT 2.50	Download
QNN_DLC	w8a16	Universal	QAIRT 2.50	Download
QNN_DLC	w8a8	Universal	QAIRT 2.50	Download
TFLITE	float	Universal	QAIRT 2.50	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® 8 Elite Gen 5 For Galaxy Mobile	0.734 ms	0 - 42 MB	NPU
SINet	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	0.878 ms	0 - 38 MB	NPU
SINet	ONNX	float	Snapdragon® X2 Elite	0.796 ms	2 - 2 MB	NPU
SINet	ONNX	float	Snapdragon® X Elite	1.843 ms	2 - 2 MB	NPU
SINet	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.173 ms	0 - 53 MB	NPU
SINet	ONNX	float	Snapdragon® 8 Gen 1 Mobile	2.111 ms	1 - 60 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	1.885 ms	1 - 5 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.765 ms	1 - 3 MB	NPU
SINet	ONNX	float	Qualcomm® QCS8450	2.111 ms	1 - 60 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	2.004 ms	1 - 5 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	1.843 ms	2 - 2 MB	NPU
SINet	ONNX	float	Qualcomm® Dragonwing™ Q-8750	0.878 ms	0 - 38 MB	NPU
SINet	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.717 ms	1 - 39 MB	NPU
SINet	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	0.856 ms	0 - 38 MB	NPU
SINet	QNN_DLC	float	Snapdragon® X2 Elite	1.009 ms	1 - 1 MB	NPU
SINet	QNN_DLC	float	Snapdragon® X Elite	1.983 ms	1 - 1 MB	NPU
SINet	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.167 ms	0 - 48 MB	NPU
SINet	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	2.163 ms	0 - 54 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	1.808 ms	1 - 5 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.753 ms	1 - 2 MB	NPU
SINet	QNN_DLC	float	Qualcomm® SA8650P	2.191 ms	0 - 36 MB	NPU
SINet	QNN_DLC	float	Qualcomm® SA8255P	2.191 ms	0 - 36 MB	NPU
SINet	QNN_DLC	float	Qualcomm® QCS8450	2.163 ms	0 - 54 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	1.991 ms	1 - 4 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	1.983 ms	1 - 1 MB	NPU
SINet	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	0.856 ms	0 - 38 MB	NPU
SINet	QNN_DLC	float	Qualcomm® SA8295P	2.46 ms	1 - 38 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.802 ms	0 - 43 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	0.984 ms	0 - 40 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® X2 Elite	1.091 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® X Elite	2.184 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	1.301 ms	0 - 52 MB	NPU
SINet	QNN_DLC	w8a16	Snapdragon® 8 Gen 1 Mobile	2.551 ms	0 - 58 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-8275	1.856 ms	0 - 4 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.945 ms	0 - 15 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® SA8650P	2.235 ms	0 - 41 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® SA8255P	2.235 ms	0 - 41 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® QCS8450	2.551 ms	0 - 58 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	2.126 ms	0 - 3 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	2.184 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	0.984 ms	0 - 40 MB	NPU
SINet	QNN_DLC	w8a16	Qualcomm® SA8295P	2.998 ms	0 - 41 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.568 ms	0 - 39 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.716 ms	0 - 42 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.823 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® X Elite	1.601 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.953 ms	0 - 50 MB	NPU
SINet	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	1.615 ms	0 - 53 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-8275	1.364 ms	0 - 3 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.401 ms	0 - 2 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® SA8650P	1.677 ms	0 - 39 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® SA8255P	1.677 ms	0 - 39 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® QCS8450	1.615 ms	0 - 53 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-9075	1.536 ms	0 - 3 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-X7181	1.601 ms	0 - 0 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-8750	0.716 ms	0 - 42 MB	NPU
SINet	QNN_DLC	w8a8	Qualcomm® SA8295P	2.017 ms	0 - 38 MB	NPU
SINet	TFLITE	float	Snapdragon® 8 Elite Gen 5 For Galaxy Mobile	0.79 ms	0 - 36 MB	NPU
SINet	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	0.93 ms	0 - 31 MB	NPU
SINet	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.23 ms	0 - 46 MB	NPU
SINet	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	2.262 ms	0 - 57 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	2.218 ms	0 - 6 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	1.812 ms	0 - 2 MB	NPU
SINet	TFLITE	float	Qualcomm® SA8650P	2.628 ms	0 - 36 MB	NPU
SINet	TFLITE	float	Qualcomm® SA8255P	2.628 ms	0 - 36 MB	NPU
SINet	TFLITE	float	Qualcomm® QCS8450	2.262 ms	0 - 57 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	2.329 ms	2 - 7 MB	NPU
SINet	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	0.93 ms	0 - 31 MB	NPU
SINet	TFLITE	float	Qualcomm® SA8295P	2.864 ms	0 - 37 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