FCN-ResNet50: Optimized for Qualcomm Devices

FCN_ResNet50 is a machine learning model that can segment images from the COCO dataset. It uses ResNet50 as a backbone.

This is based on the implementation of FCN-ResNet50 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 FCN-ResNet50 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 FCN-ResNet50 on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

• Model checkpoint: COCO_WITH_VOC_LABELS_V1
• Input resolution: 224x224
• Number of output classes: 21
• Number of parameters: 33.0M
• Model size (float): 126 MB
• Model size (w8a8): 32.2 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
FCN-ResNet50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	22.907 ms	4 - 325 MB	NPU
FCN-ResNet50	ONNX	float	Snapdragon® X2 Elite	23.315 ms	63 - 63 MB	NPU
FCN-ResNet50	ONNX	float	Snapdragon® X Elite	43.183 ms	62 - 62 MB	NPU
FCN-ResNet50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	32.653 ms	1 - 382 MB	NPU
FCN-ResNet50	ONNX	float	Qualcomm® QCS8550 (Proxy)	43.012 ms	3 - 11 MB	NPU
FCN-ResNet50	ONNX	float	Qualcomm® QCS9075	75.149 ms	3 - 9 MB	NPU
FCN-ResNet50	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	26.786 ms	1 - 297 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	7.063 ms	1 - 249 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® X2 Elite	7.115 ms	33 - 33 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® X Elite	13.857 ms	32 - 32 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	10.018 ms	1 - 280 MB	NPU
FCN-ResNet50	ONNX	w8a8	Qualcomm® QCS6490	897.613 ms	68 - 110 MB	CPU
FCN-ResNet50	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	13.52 ms	0 - 21 MB	NPU
FCN-ResNet50	ONNX	w8a8	Qualcomm® QCS9075	15.137 ms	1 - 4 MB	NPU
FCN-ResNet50	ONNX	w8a8	Qualcomm® QCM6690	839.434 ms	48 - 57 MB	CPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	8.556 ms	1 - 205 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	705.649 ms	73 - 82 MB	CPU
FCN-ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	21.759 ms	2 - 331 MB	NPU
FCN-ResNet50	QNN_DLC	float	Snapdragon® X2 Elite	22.764 ms	3 - 3 MB	NPU
FCN-ResNet50	QNN_DLC	float	Snapdragon® X Elite	44.235 ms	3 - 3 MB	NPU
FCN-ResNet50	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	33.279 ms	0 - 384 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	272.271 ms	1 - 304 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	44.391 ms	3 - 5 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® SA8775P	72.461 ms	1 - 304 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® QCS9075	78.746 ms	3 - 8 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	84.153 ms	2 - 278 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® SA7255P	272.271 ms	1 - 304 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® SA8295P	77.744 ms	0 - 217 MB	NPU
FCN-ResNet50	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	26.371 ms	0 - 318 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	7.547 ms	1 - 245 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® X2 Elite	7.831 ms	1 - 1 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® X Elite	15.211 ms	1 - 1 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	11.422 ms	1 - 265 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS6490	92.897 ms	1 - 3 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	39.517 ms	1 - 205 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	14.967 ms	1 - 3 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® SA8775P	14.691 ms	1 - 215 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS9075	16.846 ms	3 - 5 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCM6690	386.21 ms	1 - 347 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	23.809 ms	1 - 267 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® SA7255P	39.517 ms	1 - 205 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® SA8295P	22.147 ms	1 - 209 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	8.867 ms	1 - 201 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	25.958 ms	1 - 281 MB	NPU
FCN-ResNet50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	20.472 ms	0 - 355 MB	NPU
FCN-ResNet50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	33.045 ms	0 - 428 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® QCS8275 (Proxy)	272.405 ms	0 - 332 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® QCS8550 (Proxy)	44.279 ms	0 - 7 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® SA8775P	72.53 ms	0 - 331 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® QCS9075	77.969 ms	0 - 71 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® QCS8450 (Proxy)	84.925 ms	0 - 317 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® SA7255P	272.405 ms	0 - 332 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® SA8295P	77.796 ms	0 - 246 MB	NPU
FCN-ResNet50	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	27.063 ms	0 - 343 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	7.145 ms	0 - 247 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	11.04 ms	0 - 267 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS6490	94.783 ms	0 - 39 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	38.268 ms	0 - 206 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	14.405 ms	0 - 3 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® SA8775P	14.726 ms	0 - 205 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS9075	14.958 ms	0 - 35 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCM6690	450.037 ms	0 - 352 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	23.158 ms	0 - 269 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® SA7255P	38.268 ms	0 - 206 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® SA8295P	21.372 ms	0 - 210 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	8.54 ms	0 - 201 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	26.682 ms	0 - 282 MB	NPU

License

• The license for the original implementation of FCN-ResNet50 can be found here.

References

• Fully Convolutional Networks for Semantic Segmentation
• 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.