PSPNet: Optimized for Qualcomm Devices

PSPNet (Pyramid Scene Parsing Network) is a semantic segmentation model that captures global context information by applying pyramid pooling modules. It is designed to improve scene understanding by aggregating contextual features at multiple scales.

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 2.45, ONNX Runtime 1.25.0	Download
ONNX	w8a8	Universal	QAIRT 2.45, ONNX Runtime 1.25.0	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a8	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download
TFLITE	w8a8	Universal	QAIRT 2.45	Download

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

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

Model checkpoint: pspnet101_ade20k.pth
Input resolution: 1x3x473x473
Number of parameters: 65.7M
Model size (float): 251 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
PSPNet	ONNX	float	Snapdragon® X2 Elite	446.325 ms	496 - 496 MB	NPU
PSPNet	ONNX	float	Snapdragon® X Elite	694.881 ms	498 - 498 MB	NPU
PSPNet	ONNX	float	Snapdragon® 8 Gen 3 Mobile	505.453 ms	148 - 1119 MB	NPU
PSPNet	ONNX	float	Snapdragon® 8 Gen 1 Mobile	1705.101 ms	54 - 493 MB	NPU
PSPNet	ONNX	float	Qualcomm® QCS8550 (Proxy)	587.658 ms	6 - 170 MB	NPU
PSPNet	ONNX	float	Qualcomm® QCS8450	1705.101 ms	54 - 493 MB	NPU
PSPNet	ONNX	float	Snapdragon® 8 Elite Mobile	337.39 ms	256 - 1014 MB	NPU
PSPNet	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	319.761 ms	273 - 1056 MB	NPU
PSPNet	ONNX	float	Qualcomm® QCS9075	1764.712 ms	11 - 56 MB	NPU
PSPNet	ONNX	float	Qualcomm® QCS8750	337.39 ms	256 - 1014 MB	NPU
PSPNet	ONNX	float	Qualcomm® QCS7181	694.881 ms	498 - 498 MB	NPU
PSPNet	ONNX	w8a8	Snapdragon® X2 Elite	15.061 ms	212 - 212 MB	NPU
PSPNet	ONNX	w8a8	Snapdragon® X Elite	28.03 ms	148 - 148 MB	NPU
PSPNet	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	21.732 ms	71 - 476 MB	NPU
PSPNet	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	41.091 ms	69 - 462 MB	NPU
PSPNet	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	28.752 ms	28 - 150 MB	NPU
PSPNet	ONNX	w8a8	Qualcomm® QCS8450	41.091 ms	69 - 462 MB	NPU
PSPNet	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	16.636 ms	73 - 410 MB	NPU
PSPNet	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	93.782 ms	69 - 607 MB	NPU
PSPNet	ONNX	w8a8	Qualcomm® QCS9075	33.674 ms	73 - 118 MB	NPU
PSPNet	ONNX	w8a8	Snapdragon® 8 Elite Mobile	19.088 ms	72 - 340 MB	NPU
PSPNet	ONNX	w8a8	Qualcomm® QCS7790	93.782 ms	69 - 607 MB	NPU
PSPNet	ONNX	w8a8	Qualcomm® QCS8750	19.088 ms	72 - 340 MB	NPU
PSPNet	ONNX	w8a8	Qualcomm® QCS7181	28.03 ms	148 - 148 MB	NPU
PSPNet	QNN_DLC	float	Snapdragon® X2 Elite	228.176 ms	3 - 3 MB	NPU
PSPNet	QNN_DLC	float	Snapdragon® X Elite	532.47 ms	3 - 3 MB	NPU
PSPNet	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	401.396 ms	1 - 950 MB	NPU
PSPNet	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	1703.918 ms	0 - 438 MB	NPU
PSPNet	QNN_DLC	float	Qualcomm® QCS8275	1330.552 ms	2 - 727 MB	NPU
PSPNet	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	593.111 ms	3 - 5 MB	NPU
PSPNet	QNN_DLC	float	Qualcomm® QCS8450	1703.918 ms	0 - 438 MB	NPU
PSPNet	QNN_DLC	float	Snapdragon® 8 Elite Mobile	302.172 ms	0 - 724 MB	NPU
PSPNet	QNN_DLC	float	Qualcomm® SA8295P	701.273 ms	3 - 329 MB	NPU
PSPNet	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	267.5 ms	3 - 757 MB	NPU
PSPNet	QNN_DLC	float	Qualcomm® SA7255P	1330.552 ms	2 - 727 MB	NPU
PSPNet	QNN_DLC	float	Qualcomm® QCS9075	1762.009 ms	3 - 135 MB	NPU
PSPNet	QNN_DLC	float	Qualcomm® QCS8750	302.172 ms	0 - 724 MB	NPU
PSPNet	QNN_DLC	float	Qualcomm® QCS7181	532.47 ms	3 - 3 MB	NPU
PSPNet	QNN_DLC	w8a8	Snapdragon® X2 Elite	14.478 ms	1 - 1 MB	NPU
PSPNet	QNN_DLC	w8a8	Snapdragon® X Elite	26.211 ms	1 - 1 MB	NPU
PSPNet	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	20.884 ms	1 - 355 MB	NPU
PSPNet	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	37.574 ms	0 - 354 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® QCS8275	63.868 ms	1 - 260 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	28.243 ms	1 - 3 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® QCS8450	37.574 ms	0 - 354 MB	NPU
PSPNet	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	15.169 ms	1 - 312 MB	NPU
PSPNet	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	91.633 ms	1 - 436 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® QCS9075	31.773 ms	1 - 35 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® SA7255P	63.868 ms	1 - 260 MB	NPU
PSPNet	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	18.01 ms	1 - 230 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® SA8295P	38.501 ms	1 - 263 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® QCS7790	91.633 ms	1 - 436 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® QCS8750	18.01 ms	1 - 230 MB	NPU
PSPNet	QNN_DLC	w8a8	Qualcomm® QCS7181	26.211 ms	1 - 1 MB	NPU
PSPNet	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	405.838 ms	128 - 1128 MB	NPU
PSPNet	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	1506.433 ms	62 - 612 MB	NPU
PSPNet	TFLITE	float	Qualcomm® QCS8275	1327.673 ms	121 - 894 MB	NPU
PSPNet	TFLITE	float	Qualcomm® QCS8550 (Proxy)	597.704 ms	128 - 131 MB	NPU
PSPNet	TFLITE	float	Qualcomm® SA8775P	7305.793 ms	363 - 373 MB	CPU
PSPNet	TFLITE	float	Qualcomm® SA8650P	7305.793 ms	363 - 373 MB	CPU
PSPNet	TFLITE	float	Qualcomm® SA8255P	7305.793 ms	363 - 373 MB	CPU
PSPNet	TFLITE	float	Qualcomm® QCS8450	1506.433 ms	62 - 612 MB	NPU
PSPNet	TFLITE	float	Snapdragon® 8 Elite Mobile	300.002 ms	128 - 903 MB	NPU
PSPNet	TFLITE	float	Qualcomm® SA8295P	752.392 ms	128 - 557 MB	NPU
PSPNet	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	267.691 ms	128 - 929 MB	NPU
PSPNet	TFLITE	float	Qualcomm® SA7255P	1327.673 ms	121 - 894 MB	NPU
PSPNet	TFLITE	float	Qualcomm® QCS9075	1760.848 ms	68 - 339 MB	NPU
PSPNet	TFLITE	float	Qualcomm® QCS8750	300.002 ms	128 - 903 MB	NPU
PSPNet	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	20.56 ms	0 - 357 MB	NPU
PSPNet	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	37.014 ms	34 - 389 MB	NPU
PSPNet	TFLITE	w8a8	Qualcomm® QCS8275	63.825 ms	32 - 283 MB	NPU
PSPNet	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	27.568 ms	32 - 34 MB	NPU
PSPNet	TFLITE	w8a8	Qualcomm® SA8775P	2042.314 ms	93 - 104 MB	CPU
PSPNet	TFLITE	w8a8	Qualcomm® SA8650P	2042.314 ms	93 - 104 MB	CPU
PSPNet	TFLITE	w8a8	Qualcomm® SA8255P	2042.314 ms	93 - 104 MB	CPU
PSPNet	TFLITE	w8a8	Qualcomm® QCS8450	37.014 ms	34 - 389 MB	NPU
PSPNet	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	15.125 ms	32 - 316 MB	NPU
PSPNet	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	90.411 ms	32 - 447 MB	NPU
PSPNet	TFLITE	w8a8	Qualcomm® QCS9075	41.107 ms	32 - 132 MB	NPU
PSPNet	TFLITE	w8a8	Qualcomm® SA7255P	63.825 ms	32 - 283 MB	NPU
PSPNet	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	18.112 ms	30 - 268 MB	NPU
PSPNet	TFLITE	w8a8	Qualcomm® SA8295P	38.199 ms	32 - 286 MB	NPU
PSPNet	TFLITE	w8a8	Qualcomm® QCS7790	90.411 ms	32 - 447 MB	NPU
PSPNet	TFLITE	w8a8	Qualcomm® QCS8750	18.112 ms	30 - 268 MB	NPU

License

The license for the original implementation of PSPNet can be found here.

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.

Downloads last month: -; Downloads are not tracked for this model. How to track

Model tree for qualcomm/PSPNet

Finetunes

1 model