You can deploy this model and start earning money today!

qualcomm/
Unet-Segmentation

Name: Unet-Segmentation
Rating: 2.5 (149 reviews)

Views

No views yet

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

Due to licensing restrictions, we cannot distribute pre-exported model assets for this model. 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

See our repository for Unet-Segmentation on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

Input resolution: 640x1280
Model checkpoint: unet_carvana_scale1.0_epoch2
Model size (float): 118 MB
Model size (w8a8): 29.8 MB
Number of output classes: 2 (foreground / background)
Number of parameters: 31.0M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Unet-Segmentation	ONNX	float	Snapdragon® X2 Elite	74.629 ms	17 - 17 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® X Elite	141.99 ms	54 - 54 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® 8 Gen 3 Mobile	116.625 ms	1 - 523 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® 8 Gen 1 Mobile	309.327 ms	1 - 549 MB	NPU
Unet-Segmentation	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	286.15 ms	9 - 22 MB	NPU
Unet-Segmentation	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	152.561 ms	0 - 913 MB	NPU
Unet-Segmentation	ONNX	float	Qualcomm® QCS8450	309.327 ms	1 - 549 MB	NPU
Unet-Segmentation	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	249.811 ms	9 - 21 MB	NPU
Unet-Segmentation	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	141.99 ms	54 - 54 MB	NPU
Unet-Segmentation	ONNX	float	Qualcomm® Dragonwing™ Q-8750	91.325 ms	14 - 332 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® 8 Elite Mobile	91.325 ms	14 - 332 MB	NPU
Unet-Segmentation	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	65.975 ms	15 - 345 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® X2 Elite	18.766 ms	5 - 5 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® X Elite	37.742 ms	29 - 29 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	29.879 ms	6 - 339 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	65.929 ms	6 - 342 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® Dragonwing™ QCS6490	299.866 ms	3 - 8 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-8275	40.832 ms	4 - 7 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	38.105 ms	0 - 32 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® QCS8450	65.929 ms	6 - 342 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-9075	36.811 ms	4 - 7 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-X7181	37.742 ms	29 - 29 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® Dragonwing™ Q-6690	1209.561 ms	0 - 540 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® Dragonwing™ Q-7790	84.447 ms	6 - 282 MB	NPU
Unet-Segmentation	ONNX	w8a8	Qualcomm® Dragonwing™ Q-8750	24.495 ms	3 - 189 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 8 Elite Mobile	24.495 ms	3 - 189 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	16.658 ms	3 - 192 MB	NPU
Unet-Segmentation	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	84.447 ms	6 - 282 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® X2 Elite	72.005 ms	9 - 9 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® X Elite	132.84 ms	9 - 9 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	102.656 ms	9 - 527 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	274.747 ms	9 - 545 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	274.084 ms	9 - 27 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	953.506 ms	0 - 324 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	138.003 ms	9 - 12 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® SA8775P	240.415 ms	0 - 324 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® SA8650P	240.415 ms	0 - 324 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® SA8255P	240.415 ms	0 - 324 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® QCS8450	274.747 ms	9 - 545 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	244.241 ms	9 - 27 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	132.84 ms	9 - 9 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	82.7 ms	9 - 341 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® SA7255P	953.506 ms	0 - 324 MB	NPU
Unet-Segmentation	QNN_DLC	float	Qualcomm® SA8295P	274.455 ms	0 - 322 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® 8 Elite Mobile	82.7 ms	9 - 341 MB	NPU
Unet-Segmentation	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	61.547 ms	9 - 355 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® X2 Elite	18.921 ms	2 - 2 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® X Elite	35.778 ms	2 - 2 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	26.343 ms	2 - 319 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	59.235 ms	2 - 317 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS6490	289.039 ms	4 - 10 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-8275	37.053 ms	2 - 9 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-8275	121.484 ms	1 - 180 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	34.999 ms	2 - 4 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® SA8775P	32.182 ms	1 - 181 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® SA8650P	32.182 ms	1 - 181 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® SA8255P	32.182 ms	1 - 181 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® QCS8450	59.235 ms	2 - 317 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-9075	32.975 ms	1 - 6 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-X7181	35.778 ms	2 - 2 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-6690	1153.577 ms	2 - 522 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-7790	78.874 ms	2 - 267 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-8750	21.738 ms	2 - 188 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® SA7255P	121.484 ms	1 - 180 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Qualcomm® SA8295P	63.744 ms	0 - 180 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	21.738 ms	2 - 188 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	15.995 ms	2 - 200 MB	NPU
Unet-Segmentation	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	78.874 ms	2 - 267 MB	NPU
Unet-Segmentation	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	104.23 ms	5 - 577 MB	NPU
Unet-Segmentation	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	307.578 ms	8 - 588 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	273.785 ms	0 - 81 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	953.183 ms	0 - 324 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	144.387 ms	6 - 218 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® SA8775P	240.453 ms	7 - 330 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® SA8650P	240.453 ms	7 - 330 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® SA8255P	240.453 ms	7 - 330 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® QCS8450	307.578 ms	8 - 588 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	239.24 ms	0 - 80 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	81.884 ms	5 - 337 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® SA7255P	953.183 ms	0 - 324 MB	NPU
Unet-Segmentation	TFLITE	float	Qualcomm® SA8295P	274.489 ms	7 - 328 MB	NPU
Unet-Segmentation	TFLITE	float	Snapdragon® 8 Elite Mobile	81.884 ms	5 - 337 MB	NPU
Unet-Segmentation	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	61.313 ms	6 - 354 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	26.501 ms	0 - 314 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	57.707 ms	2 - 317 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® Dragonwing™ QCS6490	288.731 ms	2 - 42 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-8275	37.33 ms	2 - 39 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-8275	121.577 ms	2 - 180 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	32.772 ms	2 - 399 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® SA8775P	32.232 ms	2 - 181 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® SA8650P	32.232 ms	2 - 181 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® SA8255P	32.232 ms	2 - 181 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® QCS8450	57.707 ms	2 - 317 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-9075	32.89 ms	1 - 37 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-6690	1213.089 ms	0 - 520 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-7790	78.86 ms	2 - 271 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-8750	21.678 ms	0 - 185 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® SA7255P	121.577 ms	2 - 180 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Qualcomm® SA8295P	63.765 ms	2 - 180 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	21.678 ms	0 - 185 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	16.077 ms	1 - 199 MB	NPU
Unet-Segmentation	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	78.86 ms	2 - 271 MB	NPU

License

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

References

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.

qualcomm/Unet-Segmentation