Unet-Segmentation: Optimized for Qualcomm Devices

UNet is a machine learning model that produces a segmentation mask for an image. The most basic use case will label each pixel in the image as being in the foreground or the background. More advanced usage will assign a class label to each pixel. This version of the model was trained on the data from Kaggle's Carvana Image Masking Challenge (see https://www.kaggle.com/c/carvana-image-masking-challenge) and is used for vehicle segmentation.

This is based on the implementation of Unet-Segmentation 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 1.24.3	Download
ONNX	w8a8	Universal	QAIRT 2.42, ONNX Runtime 1.24.3	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
QNN_DLC	w8a8	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.19.1	Download
TFLITE	w8a8	Universal	QAIRT 2.43, TFLite 2.19.1	Download

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

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

• Model checkpoint: unet_carvana_scale1.0_epoch2
• Input resolution: 224x224
• Number of output classes: 2 (foreground / background)
• Number of parameters: 31.0M
• Model size (float): 118 MB
• Model size (w8a8): 29.8 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Unet-Segmentation	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	70.17 ms	5 - 328 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® X2 Elite	75.116 ms	53 - 53 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® X Elite	139.569 ms	53 - 53 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® 8 Gen 3 Mobile	108.705 ms	24 - 562 MB	NPU
Unet-Segmentation	ONNX	float	Qualcomm® QCS8550 (Proxy)	149.107 ms	0 - 56 MB	NPU
Unet-Segmentation	ONNX	float	Qualcomm® QCS9075	254.688 ms	9 - 21 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	88.809 ms	13 - 330 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	16.208 ms	6 - 190 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® X2 Elite	20.075 ms	29 - 29 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® X Elite	39.051 ms	29 - 29 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	30.608 ms	6 - 339 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® QCS6490	4666.283 ms	943 - 1000 MB	CPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	39.355 ms	2 - 7 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® QCS9075	35.609 ms	4 - 7 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® QCM6690	4122.21 ms	837 - 843 MB	CPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	24.8 ms	3 - 189 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	3883.383 ms	839 - 846 MB	CPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	62.222 ms	9 - 353 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® X2 Elite	71.768 ms	9 - 9 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® X Elite	132.421 ms	9 - 9 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	102.153 ms	9 - 545 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	953.636 ms	1 - 323 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	132.829 ms	10 - 11 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® SA8775P	240.578 ms	0 - 323 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® QCS9075	248.601 ms	9 - 27 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	281.918 ms	9 - 548 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® SA7255P	953.636 ms	1 - 323 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® SA8295P	274.496 ms	0 - 322 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	81.75 ms	9 - 340 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	15.739 ms	2 - 198 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® X2 Elite	18.818 ms	2 - 2 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® X Elite	35.661 ms	2 - 2 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	26.105 ms	2 - 320 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® QCS6490	268.196 ms	2 - 8 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	121.504 ms	1 - 180 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	34.619 ms	2 - 399 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® SA8775P	32.195 ms	1 - 180 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® QCS9075	34.661 ms	0 - 6 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® QCM6690	1198.353 ms	2 - 521 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	58.802 ms	3 - 319 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® SA7255P	121.504 ms	1 - 180 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® SA8295P	63.732 ms	1 - 181 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	21.71 ms	2 - 190 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	78.705 ms	2 - 272 MB	NPU
Unet-Segmentation	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	65.547 ms	5 - 352 MB	NPU
Unet-Segmentation	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	102.938 ms	5 - 542 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® QCS8275 (Proxy)	953.516 ms	0 - 325 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® QCS8550 (Proxy)	138.498 ms	0 - 1144 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® SA8775P	240.529 ms	0 - 323 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® QCS9075	248.107 ms	0 - 80 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® QCS8450 (Proxy)	273.884 ms	7 - 553 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® SA7255P	953.516 ms	0 - 325 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® SA8295P	274.526 ms	6 - 329 MB	NPU
Unet-Segmentation	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	81.958 ms	6 - 339 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	15.763 ms	2 - 197 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	26.05 ms	1 - 317 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® QCS6490	267.715 ms	0 - 40 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	121.562 ms	2 - 181 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	32.447 ms	2 - 4 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® SA8775P	32.245 ms	2 - 181 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® QCS9075	34.247 ms	1 - 37 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® QCM6690	1201.224 ms	0 - 518 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	61.204 ms	2 - 319 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® SA7255P	121.562 ms	2 - 181 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® SA8295P	63.789 ms	2 - 180 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	21.677 ms	1 - 188 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	78.555 ms	1 - 266 MB	NPU

License

• The license for the original implementation of Unet-Segmentation can be found here.

References

• U-Net: Convolutional Networks for Biomedical Image 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.