Inception-v3: Optimized for Qualcomm Devices

InceptionNetV3 is a machine learning model that can classify images from the Imagenet dataset. It can also be used as a backbone in building more complex models for specific use cases.

This is based on the implementation of Inception-v3 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	w8a8	Universal	QAIRT 2.45, ONNX Runtime 1.25.0	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a8	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download
TFLITE	w8a8	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit Inception-v3 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 Inception-v3 on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Model checkpoint: Imagenet
• Input resolution: 224x224
• Number of parameters: 23.9M
• Model size (float): 90.9 MB
• Model size (w8a8): 23.3 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Inception-v3	ONNX	float	Snapdragon® X2 Elite	1.141 ms	212 - 212 MB	NPU
Inception-v3	ONNX	float	Snapdragon® X Elite	2.147 ms	148 - 148 MB	NPU
Inception-v3	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.59 ms	39 - 91 MB	NPU
Inception-v3	ONNX	float	Snapdragon® 8 Gen 1 Mobile	3.963 ms	1 - 56 MB	NPU
Inception-v3	ONNX	float	Qualcomm® QCS8550 (Proxy)	2.175 ms	0 - 50 MB	NPU
Inception-v3	ONNX	float	Qualcomm® QCS8450	3.963 ms	1 - 56 MB	NPU
Inception-v3	ONNX	float	Snapdragon® 8 Elite Mobile	1.34 ms	0 - 43 MB	NPU
Inception-v3	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.07 ms	1 - 37 MB	NPU
Inception-v3	ONNX	float	Qualcomm® QCS9075	3.709 ms	1 - 46 MB	NPU
Inception-v3	ONNX	float	Qualcomm® QCS8750	1.34 ms	0 - 43 MB	NPU
Inception-v3	ONNX	float	Qualcomm® QCS7181	2.147 ms	148 - 148 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® X2 Elite	0.383 ms	181 - 181 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® X Elite	0.801 ms	149 - 149 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.635 ms	0 - 92 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	1.237 ms	0 - 92 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS6490	3.451 ms	0 - 46 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	0.852 ms	0 - 4 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS8450	1.237 ms	0 - 92 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.431 ms	0 - 57 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	1.367 ms	0 - 59 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCM6690	10.734 ms	0 - 176 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS9075	0.888 ms	0 - 45 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® 8 Elite Mobile	0.538 ms	0 - 63 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS7790	1.367 ms	0 - 59 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS8750	0.538 ms	0 - 63 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS7181	0.801 ms	149 - 149 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® X2 Elite	1.194 ms	1 - 1 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® X Elite	2.099 ms	1 - 1 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.511 ms	0 - 59 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	3.731 ms	0 - 59 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS8275	13.482 ms	1 - 35 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	1.978 ms	0 - 24 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS8450	3.731 ms	0 - 59 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® 8 Elite Mobile	1.274 ms	0 - 36 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® SA8295P	4.158 ms	0 - 33 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	0.998 ms	1 - 40 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® SA7255P	13.482 ms	1 - 35 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS9075	3.432 ms	3 - 6 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS8750	1.274 ms	0 - 36 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS7181	2.099 ms	1 - 1 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.458 ms	0 - 0 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® X Elite	0.866 ms	0 - 0 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.593 ms	0 - 80 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	1.19 ms	0 - 82 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS6490	3.524 ms	2 - 4 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS8275	2.432 ms	0 - 53 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	0.809 ms	0 - 2 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS8450	1.19 ms	0 - 82 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.412 ms	0 - 57 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.302 ms	0 - 55 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCM6690	10.581 ms	2 - 170 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS9075	0.828 ms	2 - 4 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® SA7255P	2.432 ms	0 - 53 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	0.521 ms	0 - 53 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® SA8295P	1.592 ms	0 - 52 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS7790	1.302 ms	0 - 55 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS8750	0.521 ms	0 - 53 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS7181	0.866 ms	0 - 0 MB	NPU
Inception-v3	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.5 ms	0 - 117 MB	NPU
Inception-v3	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	3.731 ms	0 - 119 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® QCS8275	13.491 ms	0 - 66 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® QCS8550 (Proxy)	2.03 ms	0 - 2 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® SA8775P	33.495 ms	0 - 30 MB	GPU
Inception-v3	TFLITE	float	Qualcomm® SA8650P	33.495 ms	0 - 30 MB	GPU
Inception-v3	TFLITE	float	Qualcomm® SA8255P	33.495 ms	0 - 30 MB	GPU
Inception-v3	TFLITE	float	Qualcomm® QCS8450	3.731 ms	0 - 119 MB	NPU
Inception-v3	TFLITE	float	Snapdragon® 8 Elite Mobile	1.264 ms	0 - 73 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® SA8295P	4.176 ms	0 - 64 MB	NPU
Inception-v3	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	0.993 ms	0 - 68 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® SA7255P	13.491 ms	0 - 66 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® QCS9075	3.448 ms	0 - 49 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® QCS8750	1.264 ms	0 - 73 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.637 ms	0 - 87 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	1.293 ms	0 - 85 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS6490	3.672 ms	0 - 25 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS8275	2.508 ms	0 - 51 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	0.91 ms	0 - 27 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® SA8775P	34.209 ms	1 - 34 MB	GPU
Inception-v3	TFLITE	w8a8	Qualcomm® SA8650P	34.209 ms	1 - 34 MB	GPU
Inception-v3	TFLITE	w8a8	Qualcomm® SA8255P	34.209 ms	1 - 34 MB	GPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS8450	1.293 ms	0 - 85 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.452 ms	0 - 55 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.4 ms	0 - 53 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCM6690	10.715 ms	0 - 168 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS9075	0.994 ms	0 - 25 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® SA7255P	2.508 ms	0 - 51 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	0.545 ms	0 - 51 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® SA8295P	1.835 ms	0 - 49 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS7790	1.4 ms	0 - 53 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS8750	0.545 ms	0 - 51 MB	NPU

License

• The license for the original implementation of Inception-v3 can be found here.

References

• Rethinking the Inception Architecture for Computer Vision
• 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.