PidNet: Optimized for Qualcomm Devices

PIDNet (Proportional-Integral-Derivative Network) is a real-time semantic segmentation model based on PID controllers

This is based on the implementation of PidNet 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 PidNet 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 PidNet on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

Model checkpoint: PIDNet_S_Cityscapes_val.pt
Inference latency: RealTime
Input resolution: 1024x2048
Number of output classes: 19
Number of parameters: 8.06M
Model size (float): 29.1 MB
Model size (w8a8): 8.02 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
PidNet	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	12.164 ms	31 - 299 MB	NPU
PidNet	ONNX	float	Snapdragon® X2 Elite	12.927 ms	22 - 22 MB	NPU
PidNet	ONNX	float	Snapdragon® X Elite	32.82 ms	24 - 24 MB	NPU
PidNet	ONNX	float	Snapdragon® 8 Gen 3 Mobile	22.949 ms	32 - 349 MB	NPU
PidNet	ONNX	float	Qualcomm® QCS8550 (Proxy)	32.516 ms	24 - 26 MB	NPU
PidNet	ONNX	float	Qualcomm® QCS9075	47.406 ms	24 - 51 MB	NPU
PidNet	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	16.324 ms	6 - 225 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	44.924 ms	94 - 327 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® X2 Elite	45.18 ms	133 - 133 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® X Elite	62.602 ms	131 - 131 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	45.704 ms	107 - 381 MB	NPU
PidNet	ONNX	w8a8	Qualcomm® QCS6490	388.418 ms	197 - 215 MB	CPU
PidNet	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	59.488 ms	97 - 104 MB	NPU
PidNet	ONNX	w8a8	Qualcomm® QCS9075	64.586 ms	100 - 103 MB	NPU
PidNet	ONNX	w8a8	Qualcomm® QCM6690	346.538 ms	129 - 138 MB	CPU
PidNet	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	44.51 ms	101 - 319 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	327.615 ms	134 - 143 MB	CPU
PidNet	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	11.942 ms	24 - 296 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® X2 Elite	13.424 ms	24 - 24 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® X Elite	38.874 ms	24 - 24 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	25.702 ms	18 - 337 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	118.298 ms	24 - 246 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	37.675 ms	24 - 26 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® SA8775P	46.714 ms	24 - 247 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® QCS9075	61.699 ms	24 - 52 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	76.751 ms	5 - 334 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® SA7255P	118.298 ms	24 - 246 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® SA8295P	51.765 ms	24 - 260 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	18.276 ms	21 - 266 MB	NPU
PidNet	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	47.613 ms	6 - 270 MB	NPU
PidNet	QNN_DLC	w8a8	Snapdragon® X2 Elite	48.386 ms	6 - 6 MB	NPU
PidNet	QNN_DLC	w8a8	Snapdragon® X Elite	60.352 ms	6 - 6 MB	NPU
PidNet	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	43.153 ms	6 - 272 MB	NPU
PidNet	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	111.214 ms	6 - 218 MB	NPU
PidNet	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	57.898 ms	6 - 8 MB	NPU
PidNet	QNN_DLC	w8a8	Qualcomm® SA8775P	58.479 ms	6 - 218 MB	NPU
PidNet	QNN_DLC	w8a8	Qualcomm® QCS9075	61.958 ms	6 - 14 MB	NPU
PidNet	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	63.925 ms	6 - 274 MB	NPU
PidNet	QNN_DLC	w8a8	Qualcomm® SA7255P	111.214 ms	6 - 218 MB	NPU
PidNet	QNN_DLC	w8a8	Qualcomm® SA8295P	66.474 ms	6 - 220 MB	NPU
PidNet	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	41.463 ms	6 - 241 MB	NPU
PidNet	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	12.705 ms	2 - 272 MB	NPU
PidNet	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	24.882 ms	1 - 334 MB	NPU
PidNet	TFLITE	float	Qualcomm® QCS8275 (Proxy)	117.12 ms	3 - 232 MB	NPU
PidNet	TFLITE	float	Qualcomm® QCS8550 (Proxy)	35.77 ms	0 - 2 MB	NPU
PidNet	TFLITE	float	Qualcomm® SA8775P	45.274 ms	2 - 233 MB	NPU
PidNet	TFLITE	float	Qualcomm® QCS9075	59.677 ms	0 - 45 MB	NPU
PidNet	TFLITE	float	Qualcomm® QCS8450 (Proxy)	76.026 ms	2 - 350 MB	NPU
PidNet	TFLITE	float	Qualcomm® SA7255P	117.12 ms	3 - 232 MB	NPU
PidNet	TFLITE	float	Qualcomm® SA8295P	50.659 ms	2 - 250 MB	NPU
PidNet	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	18.96 ms	1 - 258 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	40.653 ms	1 - 265 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	38.135 ms	0 - 266 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS6490	210.57 ms	2 - 72 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	98.263 ms	1 - 212 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	50.321 ms	1 - 2 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® SA8775P	50.963 ms	1 - 212 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS9075	53.14 ms	0 - 16 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCM6690	238.856 ms	3 - 198 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	58.319 ms	0 - 267 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® SA7255P	98.263 ms	1 - 212 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® SA8295P	57.97 ms	1 - 215 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	38.265 ms	1 - 237 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	66.782 ms	3 - 222 MB	NPU

License

The license for the original implementation of PidNet 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.

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

Paper for qualcomm/PidNet

PIDNet: A Real-time Semantic Segmentation Network Inspired by PID Controllers

Paper • 2206.02066 • Published Jun 4, 2022