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 2.42, ONNX Runtime 1.24.1	Download
ONNX	w8a8	Universal	QAIRT 2.42, ONNX Runtime 1.24.1	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
QNN_DLC	w8a8	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.17.0	Download
TFLITE	w8a8	Universal	QAIRT 2.43, TFLite 2.17.0	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® 8 Elite Gen 5 Mobile	0.808 ms	1 - 39 MB	NPU
Inception-v3	ONNX	float	Snapdragon® X2 Elite	0.7 ms	46 - 46 MB	NPU
Inception-v3	ONNX	float	Snapdragon® X Elite	1.519 ms	46 - 46 MB	NPU
Inception-v3	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.05 ms	0 - 57 MB	NPU
Inception-v3	ONNX	float	Qualcomm® QCS8550 (Proxy)	1.491 ms	0 - 227 MB	NPU
Inception-v3	ONNX	float	Qualcomm® QCS9075	2.183 ms	1 - 4 MB	NPU
Inception-v3	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	0.925 ms	0 - 38 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.41 ms	0 - 57 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® X2 Elite	0.282 ms	23 - 23 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® X Elite	0.719 ms	23 - 23 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	0.504 ms	0 - 90 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS6490	24.943 ms	8 - 37 MB	CPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	0.66 ms	0 - 27 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCS9075	0.738 ms	0 - 3 MB	NPU
Inception-v3	ONNX	w8a8	Qualcomm® QCM6690	17.977 ms	10 - 19 MB	CPU
Inception-v3	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.432 ms	0 - 52 MB	NPU
Inception-v3	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	13.904 ms	8 - 19 MB	CPU
Inception-v3	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	0.763 ms	1 - 37 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® X2 Elite	0.802 ms	1 - 1 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® X Elite	1.43 ms	1 - 1 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.013 ms	0 - 57 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	7.849 ms	1 - 34 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	1.367 ms	1 - 145 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® SA8775P	2.241 ms	1 - 35 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS9075	2.052 ms	1 - 3 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	2.346 ms	0 - 55 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® SA7255P	7.849 ms	1 - 34 MB	NPU
Inception-v3	QNN_DLC	float	Qualcomm® SA8295P	2.689 ms	0 - 30 MB	NPU
Inception-v3	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	0.878 ms	0 - 34 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.371 ms	0 - 51 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.356 ms	0 - 0 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® X Elite	0.68 ms	0 - 0 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.482 ms	0 - 76 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS6490	2.241 ms	2 - 4 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	1.565 ms	0 - 47 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	0.608 ms	0 - 3 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® SA8775P	0.784 ms	0 - 51 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS9075	0.629 ms	0 - 2 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCM6690	5.764 ms	0 - 50 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	0.895 ms	0 - 74 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® SA7255P	1.565 ms	0 - 47 MB	NPU
Inception-v3	QNN_DLC	w8a8	Qualcomm® SA8295P	1.141 ms	0 - 46 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.404 ms	0 - 47 MB	NPU
Inception-v3	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	0.978 ms	0 - 50 MB	NPU
Inception-v3	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	0.761 ms	0 - 76 MB	NPU
Inception-v3	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.018 ms	0 - 123 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® QCS8275 (Proxy)	7.844 ms	0 - 73 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® QCS8550 (Proxy)	1.378 ms	0 - 197 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® SA8775P	2.258 ms	0 - 75 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® QCS9075	2.059 ms	0 - 48 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® QCS8450 (Proxy)	2.362 ms	0 - 126 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® SA7255P	7.844 ms	0 - 73 MB	NPU
Inception-v3	TFLITE	float	Qualcomm® SA8295P	2.683 ms	0 - 70 MB	NPU
Inception-v3	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	0.889 ms	0 - 77 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.373 ms	0 - 49 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.478 ms	0 - 74 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS6490	2.244 ms	0 - 25 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	1.629 ms	0 - 46 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	0.644 ms	0 - 1 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® SA8775P	0.884 ms	0 - 50 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS9075	0.705 ms	0 - 25 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCM6690	5.844 ms	0 - 50 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	0.939 ms	0 - 75 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® SA7255P	1.629 ms	0 - 46 MB	NPU
Inception-v3	TFLITE	w8a8	Qualcomm® SA8295P	1.26 ms	0 - 45 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.413 ms	0 - 45 MB	NPU
Inception-v3	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.011 ms	0 - 49 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.