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® X2 Elite	23.447 ms	63 - 63 MB	NPU
FCN-ResNet50	ONNX	float	Snapdragon® X Elite	43.269 ms	62 - 62 MB	NPU
FCN-ResNet50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	32.581 ms	0 - 378 MB	NPU
FCN-ResNet50	ONNX	float	Qualcomm® QCS8550 (Proxy)	43.247 ms	3 - 6 MB	NPU
FCN-ResNet50	ONNX	float	Qualcomm® QCS9075	75.296 ms	3 - 9 MB	NPU
FCN-ResNet50	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	26.716 ms	1 - 309 MB	NPU
FCN-ResNet50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	22.542 ms	4 - 325 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® X2 Elite	7.109 ms	33 - 33 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® X Elite	13.851 ms	32 - 32 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	10.035 ms	0 - 279 MB	NPU
FCN-ResNet50	ONNX	w8a8	Qualcomm® QCS6490	912.318 ms	66 - 110 MB	CPU
FCN-ResNet50	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	13.534 ms	1 - 4 MB	NPU
FCN-ResNet50	ONNX	w8a8	Qualcomm® QCS9075	15.135 ms	1 - 4 MB	NPU
FCN-ResNet50	ONNX	w8a8	Qualcomm® QCM6690	835.668 ms	66 - 74 MB	CPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	8.613 ms	1 - 203 MB	NPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	703.386 ms	69 - 78 MB	CPU
FCN-ResNet50	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	7.057 ms	0 - 248 MB	NPU
FCN-ResNet50	QNN_DLC	float	Snapdragon® X2 Elite	22.759 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	33.595 ms	3 - 387 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	272.3 ms	1 - 304 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	44.873 ms	3 - 33 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® SA8775P	72.455 ms	1 - 305 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® QCS9075	78.99 ms	5 - 10 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	84.255 ms	3 - 280 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® SA7255P	272.3 ms	1 - 304 MB	NPU
FCN-ResNet50	QNN_DLC	float	Qualcomm® SA8295P	77.735 ms	0 - 218 MB	NPU
FCN-ResNet50	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	26.506 ms	0 - 319 MB	NPU
FCN-ResNet50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	21.795 ms	2 - 330 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® X2 Elite	7.88 ms	1 - 1 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® X Elite	15.169 ms	1 - 1 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	11.475 ms	1 - 268 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS6490	93.25 ms	3 - 5 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	39.566 ms	1 - 205 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	14.992 ms	1 - 3 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® SA8775P	64.676 ms	1 - 215 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS9075	16.846 ms	1 - 3 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCM6690	416.057 ms	1 - 347 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	24.324 ms	1 - 266 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® SA7255P	39.566 ms	1 - 205 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Qualcomm® SA8295P	22.131 ms	1 - 209 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	8.868 ms	0 - 200 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	25.738 ms	1 - 280 MB	NPU
FCN-ResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	7.554 ms	1 - 245 MB	NPU
FCN-ResNet50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	33.053 ms	0 - 427 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® QCS8275 (Proxy)	272.32 ms	0 - 332 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® QCS8550 (Proxy)	44.582 ms	0 - 3 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® SA8775P	72.525 ms	0 - 331 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® QCS9075	77.893 ms	0 - 71 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® QCS8450 (Proxy)	84.829 ms	0 - 318 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® SA7255P	272.32 ms	0 - 332 MB	NPU
FCN-ResNet50	TFLITE	float	Qualcomm® SA8295P	77.792 ms	0 - 246 MB	NPU
FCN-ResNet50	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	27.206 ms	0 - 343 MB	NPU
FCN-ResNet50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	20.506 ms	0 - 351 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	11.056 ms	0 - 270 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS6490	94.88 ms	0 - 39 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	38.248 ms	0 - 206 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	14.338 ms	0 - 2 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® SA8775P	62.896 ms	0 - 205 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS9075	15.34 ms	0 - 35 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCM6690	426.83 ms	0 - 353 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	23.255 ms	0 - 267 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® SA7255P	38.248 ms	0 - 206 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Qualcomm® SA8295P	21.364 ms	0 - 209 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	8.551 ms	0 - 197 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	26.585 ms	0 - 282 MB	NPU
FCN-ResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	7.164 ms	0 - 246 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.