DenseNet-121: Optimized for Qualcomm Devices

Densenet 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 DenseNet-121 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	w8a16	Universal	QAIRT 2.42, ONNX Runtime 1.24.1	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
QNN_DLC	w8a16	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 DenseNet-121 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 DenseNet-121 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: 7.99M
• Model size (float): 30.5 MB
• Model size (w8a16): 8.72 MB
• Model size (w8a8): 8.30 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
DenseNet-121	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	0.682 ms	0 - 46 MB	NPU
DenseNet-121	ONNX	float	Snapdragon® X2 Elite	0.743 ms	16 - 16 MB	NPU
DenseNet-121	ONNX	float	Snapdragon® X Elite	1.733 ms	15 - 15 MB	NPU
DenseNet-121	ONNX	float	Snapdragon® 8 Gen 3 Mobile	1.071 ms	0 - 93 MB	NPU
DenseNet-121	ONNX	float	Qualcomm® QCS8550 (Proxy)	1.524 ms	0 - 26 MB	NPU
DenseNet-121	ONNX	float	Qualcomm® QCS9075	2.611 ms	1 - 3 MB	NPU
DenseNet-121	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	0.784 ms	0 - 46 MB	NPU
DenseNet-121	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	1.117 ms	0 - 79 MB	NPU
DenseNet-121	ONNX	w8a16	Snapdragon® X2 Elite	1.236 ms	27 - 27 MB	NPU
DenseNet-121	ONNX	w8a16	Snapdragon® X Elite	2.579 ms	25 - 25 MB	NPU
DenseNet-121	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	1.719 ms	0 - 106 MB	NPU
DenseNet-121	ONNX	w8a16	Qualcomm® QCS6490	244.564 ms	32 - 35 MB	CPU
DenseNet-121	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	2.342 ms	0 - 34 MB	NPU
DenseNet-121	ONNX	w8a16	Qualcomm® QCS9075	2.928 ms	0 - 3 MB	NPU
DenseNet-121	ONNX	w8a16	Qualcomm® QCM6690	114.861 ms	22 - 32 MB	CPU
DenseNet-121	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	1.325 ms	0 - 77 MB	NPU
DenseNet-121	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	103.712 ms	30 - 42 MB	CPU
DenseNet-121	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	0.708 ms	1 - 43 MB	NPU
DenseNet-121	QNN_DLC	float	Snapdragon® X2 Elite	0.94 ms	1 - 1 MB	NPU
DenseNet-121	QNN_DLC	float	Snapdragon® X Elite	2.041 ms	1 - 1 MB	NPU
DenseNet-121	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.198 ms	0 - 79 MB	NPU
DenseNet-121	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	7.766 ms	1 - 39 MB	NPU
DenseNet-121	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	1.795 ms	1 - 2 MB	NPU
DenseNet-121	QNN_DLC	float	Qualcomm® SA8775P	2.694 ms	1 - 43 MB	NPU
DenseNet-121	QNN_DLC	float	Qualcomm® QCS9075	2.768 ms	3 - 5 MB	NPU
DenseNet-121	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	3.283 ms	0 - 77 MB	NPU
DenseNet-121	QNN_DLC	float	Qualcomm® SA7255P	7.766 ms	1 - 39 MB	NPU
DenseNet-121	QNN_DLC	float	Qualcomm® SA8295P	3.046 ms	0 - 36 MB	NPU
DenseNet-121	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	0.865 ms	1 - 43 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	1.215 ms	0 - 72 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Snapdragon® X2 Elite	1.571 ms	0 - 0 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Snapdragon® X Elite	3.445 ms	0 - 0 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	2.147 ms	0 - 88 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Qualcomm® QCS8275 (Proxy)	6.274 ms	0 - 62 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Qualcomm® QCS8550 (Proxy)	3.126 ms	0 - 2 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Qualcomm® SA8775P	3.498 ms	0 - 65 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Qualcomm® QCS9075	3.41 ms	2 - 4 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Qualcomm® QCS8450 (Proxy)	5.009 ms	0 - 86 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Qualcomm® SA7255P	6.274 ms	0 - 62 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Qualcomm® SA8295P	5.599 ms	0 - 60 MB	NPU
DenseNet-121	QNN_DLC	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	1.76 ms	0 - 60 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.901 ms	0 - 65 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Snapdragon® X2 Elite	1.151 ms	0 - 0 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Snapdragon® X Elite	2.639 ms	0 - 0 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	1.695 ms	0 - 77 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	4.915 ms	0 - 57 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	2.376 ms	0 - 2 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Qualcomm® SA8775P	2.694 ms	0 - 60 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Qualcomm® QCS9075	2.573 ms	2 - 4 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	2.955 ms	0 - 80 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Qualcomm® SA7255P	4.915 ms	0 - 57 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Qualcomm® SA8295P	3.464 ms	0 - 55 MB	NPU
DenseNet-121	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	1.051 ms	0 - 59 MB	NPU
DenseNet-121	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	0.708 ms	0 - 59 MB	NPU
DenseNet-121	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.199 ms	0 - 97 MB	NPU
DenseNet-121	TFLITE	float	Qualcomm® QCS8275 (Proxy)	7.75 ms	0 - 55 MB	NPU
DenseNet-121	TFLITE	float	Qualcomm® QCS8550 (Proxy)	1.794 ms	0 - 6 MB	NPU
DenseNet-121	TFLITE	float	Qualcomm® SA8775P	2.715 ms	0 - 57 MB	NPU
DenseNet-121	TFLITE	float	Qualcomm® QCS9075	2.831 ms	0 - 18 MB	NPU
DenseNet-121	TFLITE	float	Qualcomm® QCS8450 (Proxy)	3.292 ms	0 - 92 MB	NPU
DenseNet-121	TFLITE	float	Qualcomm® SA7255P	7.75 ms	0 - 55 MB	NPU
DenseNet-121	TFLITE	float	Qualcomm® SA8295P	3.055 ms	0 - 53 MB	NPU
DenseNet-121	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	0.857 ms	0 - 60 MB	NPU
DenseNet-121	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.858 ms	0 - 65 MB	NPU
DenseNet-121	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	1.583 ms	0 - 79 MB	NPU
DenseNet-121	TFLITE	w8a8	Qualcomm® QCS6490	62.588 ms	6 - 33 MB	NPU
DenseNet-121	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	4.572 ms	0 - 58 MB	NPU
DenseNet-121	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	2.231 ms	0 - 3 MB	NPU
DenseNet-121	TFLITE	w8a8	Qualcomm® SA8775P	2.585 ms	0 - 60 MB	NPU
DenseNet-121	TFLITE	w8a8	Qualcomm® QCS9075	2.428 ms	0 - 11 MB	NPU
DenseNet-121	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	2.646 ms	0 - 81 MB	NPU
DenseNet-121	TFLITE	w8a8	Qualcomm® SA7255P	4.572 ms	0 - 58 MB	NPU
DenseNet-121	TFLITE	w8a8	Qualcomm® SA8295P	3.311 ms	0 - 57 MB	NPU
DenseNet-121	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	1.005 ms	0 - 56 MB	NPU
DenseNet-121	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	26.662 ms	6 - 171 MB	NPU

License

• The license for the original implementation of DenseNet-121 can be found here.

References

• Densely Connected Convolutional Networks
• 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.