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 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download
TFLITE	w8a8	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit 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	12.126 ms	30 - 297 MB	NPU
PidNet	ONNX	float	Snapdragon® 8 Elite Mobile	16.084 ms	6 - 220 MB	NPU
PidNet	ONNX	float	Snapdragon® X2 Elite	12.94 ms	22 - 22 MB	NPU
PidNet	ONNX	float	Snapdragon® X Elite	32.643 ms	24 - 24 MB	NPU
PidNet	ONNX	float	Snapdragon® X Elite	32.643 ms	24 - 24 MB	NPU
PidNet	ONNX	float	Snapdragon® 8 Gen 3 Mobile	22.893 ms	30 - 348 MB	NPU
PidNet	ONNX	float	Qualcomm® QCS8550 (Proxy)	33.384 ms	24 - 27 MB	NPU
PidNet	ONNX	float	Qualcomm® QCS9075	47.268 ms	24 - 51 MB	NPU
PidNet	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	16.084 ms	6 - 220 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	44.797 ms	98 - 319 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® 8 Elite Mobile	44.138 ms	103 - 314 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® X2 Elite	45.924 ms	134 - 134 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® X Elite	84.746 ms	133 - 133 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® X Elite	84.746 ms	133 - 133 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	45.517 ms	87 - 348 MB	NPU
PidNet	ONNX	w8a8	Qualcomm® QCS6490	387.67 ms	197 - 216 MB	CPU
PidNet	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	59.563 ms	103 - 107 MB	NPU
PidNet	ONNX	w8a8	Qualcomm® QCS9075	64.678 ms	104 - 106 MB	NPU
PidNet	ONNX	w8a8	Qualcomm® QCM6690	345.104 ms	196 - 206 MB	CPU
PidNet	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	44.138 ms	103 - 314 MB	NPU
PidNet	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	327.875 ms	136 - 146 MB	CPU
PidNet	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	327.875 ms	136 - 146 MB	CPU
PidNet	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	11.999 ms	24 - 293 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® 8 Elite Mobile	18.241 ms	15 - 260 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® X2 Elite	13.515 ms	24 - 24 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® X Elite	38.799 ms	24 - 24 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® X Elite	38.799 ms	24 - 24 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	25.666 ms	11 - 315 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	117.778 ms	24 - 242 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	37.338 ms	24 - 68 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® SA8775P	46.595 ms	24 - 243 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® SA8775P	46.595 ms	24 - 243 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® SA8775P	46.595 ms	24 - 243 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® QCS9075	60.97 ms	24 - 52 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	75.971 ms	5 - 313 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® SA7255P	117.778 ms	24 - 242 MB	NPU
PidNet	QNN_DLC	float	Qualcomm® SA8295P	51.965 ms	24 - 252 MB	NPU
PidNet	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	18.241 ms	15 - 260 MB	NPU
PidNet	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	12.705 ms	2 - 274 MB	NPU
PidNet	TFLITE	float	Snapdragon® 8 Elite Mobile	18.933 ms	2 - 256 MB	NPU
PidNet	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	25.047 ms	2 - 327 MB	NPU
PidNet	TFLITE	float	Qualcomm® QCS8275 (Proxy)	116.623 ms	3 - 232 MB	NPU
PidNet	TFLITE	float	Qualcomm® QCS8550 (Proxy)	36.533 ms	2 - 6 MB	NPU
PidNet	TFLITE	float	Qualcomm® SA8775P	45.286 ms	3 - 232 MB	NPU
PidNet	TFLITE	float	Qualcomm® SA8775P	45.286 ms	3 - 232 MB	NPU
PidNet	TFLITE	float	Qualcomm® SA8775P	45.286 ms	3 - 232 MB	NPU
PidNet	TFLITE	float	Qualcomm® QCS9075	59.79 ms	0 - 45 MB	NPU
PidNet	TFLITE	float	Qualcomm® QCS8450 (Proxy)	74.478 ms	2 - 329 MB	NPU
PidNet	TFLITE	float	Qualcomm® SA7255P	116.623 ms	3 - 232 MB	NPU
PidNet	TFLITE	float	Qualcomm® SA8295P	51.038 ms	2 - 244 MB	NPU
PidNet	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	18.933 ms	2 - 256 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	22.133 ms	1 - 263 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	69.547 ms	1 - 234 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	38.027 ms	0 - 262 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS6490	204.74 ms	2 - 73 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	98.434 ms	1 - 213 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	50.224 ms	1 - 4 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® SA8775P	51.017 ms	1 - 214 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® SA8775P	51.017 ms	1 - 214 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® SA8775P	51.017 ms	1 - 214 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS9075	53.009 ms	1 - 17 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCM6690	233.288 ms	2 - 233 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	58.463 ms	1 - 266 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® SA7255P	98.434 ms	1 - 213 MB	NPU
PidNet	TFLITE	w8a8	Qualcomm® SA8295P	58.036 ms	1 - 218 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	69.547 ms	1 - 234 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	66.332 ms	3 - 221 MB	NPU
PidNet	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	66.332 ms	3 - 221 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