Hugging Face's logo

qualcomm
/
PidNet

Image Segmentation
PyTorch
real_time
android
Model card Files
xet
 Community

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 2.45, ONNX Runtime 1.27.1 Download
ONNX w8a8 Universal QAIRT 2.45, ONNX Runtime 1.27.1 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 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:

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

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
PidNet ONNX float Snapdragon® X2 Elite 13.731 ms 24 - 24 MB NPU
PidNet ONNX float Snapdragon® X Elite 32.116 ms 24 - 24 MB NPU
PidNet ONNX float Snapdragon® 8 Gen 3 Mobile 22.866 ms 30 - 333 MB NPU
PidNet ONNX float Snapdragon® 8 Gen 1 Mobile 55.285 ms 3 - 307 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ IQ-8275 40.513 ms 24 - 51 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 32.114 ms 24 - 396 MB NPU
PidNet ONNX float Qualcomm® QCS8450 55.285 ms 3 - 307 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ IQ-9075 41.981 ms 24 - 51 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ IQ-X7181 32.116 ms 24 - 24 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ Q-8750 17.7 ms 5 - 218 MB NPU
PidNet ONNX float Snapdragon® 8 Elite Mobile 17.7 ms 5 - 218 MB NPU
PidNet ONNX float Snapdragon® 8 Elite Gen 5 Mobile 13.15 ms 5 - 255 MB NPU
PidNet ONNX w8a8 Snapdragon® X2 Elite 26.09 ms 7 - 7 MB NPU
PidNet ONNX w8a8 Snapdragon® X Elite 55.018 ms 7 - 7 MB NPU
PidNet ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 39.048 ms 7 - 272 MB NPU
PidNet ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 57.501 ms 8 - 271 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 158.46 ms 6 - 9 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ IQ-8275 51.33 ms 6 - 10 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 52.484 ms 6 - 139 MB NPU
PidNet ONNX w8a8 Qualcomm® QCS8450 57.501 ms 8 - 271 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 54.247 ms 6 - 9 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 55.018 ms 7 - 7 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 213.628 ms 5 - 224 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 62.68 ms 7 - 235 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 45.328 ms 1 - 218 MB NPU
PidNet ONNX w8a8 Snapdragon® 8 Elite Mobile 45.328 ms 1 - 218 MB NPU
PidNet ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 25.767 ms 2 - 235 MB NPU
PidNet ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 62.68 ms 7 - 235 MB NPU
PidNet QNN_DLC float Snapdragon® X2 Elite 13.642 ms 24 - 24 MB NPU
PidNet QNN_DLC float Snapdragon® X Elite 38.789 ms 24 - 24 MB NPU
PidNet QNN_DLC float Snapdragon® 8 Gen 3 Mobile 25.489 ms 22 - 331 MB NPU
PidNet QNN_DLC float Snapdragon® 8 Gen 1 Mobile 76.05 ms 5 - 310 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 44.518 ms 24 - 53 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 117.763 ms 24 - 246 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 37.579 ms 24 - 26 MB NPU
PidNet QNN_DLC float Qualcomm® SA8775P 46.652 ms 24 - 247 MB NPU
PidNet QNN_DLC float Qualcomm® SA8650P 46.652 ms 24 - 247 MB NPU
PidNet QNN_DLC float Qualcomm® SA8255P 46.652 ms 24 - 247 MB NPU
PidNet QNN_DLC float Qualcomm® QCS8450 76.05 ms 5 - 310 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 47.37 ms 24 - 52 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 38.789 ms 24 - 24 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ Q-8750 18.277 ms 16 - 261 MB NPU
PidNet QNN_DLC float Qualcomm® SA7255P 117.763 ms 24 - 246 MB NPU
PidNet QNN_DLC float Qualcomm® SA8295P 51.969 ms 24 - 252 MB NPU
PidNet QNN_DLC float Snapdragon® 8 Elite Mobile 18.277 ms 16 - 261 MB NPU
PidNet QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 12.058 ms 16 - 287 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® X2 Elite 25.237 ms 6 - 6 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® X Elite 60.571 ms 6 - 6 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 42.908 ms 6 - 267 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 64.343 ms 6 - 270 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 56.484 ms 2 - 11 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 111.458 ms 6 - 218 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 57.651 ms 6 - 8 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA8775P 58.53 ms 6 - 219 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA8650P 58.53 ms 6 - 219 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA8255P 58.53 ms 6 - 219 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® QCS8450 64.343 ms 6 - 270 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 59.003 ms 6 - 14 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 60.571 ms 6 - 6 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 43.854 ms 0 - 238 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA7255P 111.458 ms 6 - 218 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA8295P 66.47 ms 6 - 223 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 43.854 ms 0 - 238 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 24.592 ms 6 - 273 MB NPU
PidNet TFLITE float Snapdragon® 8 Gen 3 Mobile 25.144 ms 1 - 323 MB NPU
PidNet TFLITE float Snapdragon® 8 Gen 1 Mobile 74.579 ms 0 - 325 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ IQ-8275 43.786 ms 0 - 46 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ IQ-8275 116.663 ms 0 - 228 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 36.151 ms 2 - 5 MB NPU
PidNet TFLITE float Qualcomm® SA8775P 45.289 ms 3 - 233 MB NPU
PidNet TFLITE float Qualcomm® SA8650P 45.289 ms 3 - 233 MB NPU
PidNet TFLITE float Qualcomm® SA8255P 45.289 ms 3 - 233 MB NPU
PidNet TFLITE float Qualcomm® QCS8450 74.579 ms 0 - 325 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ IQ-9075 46.749 ms 0 - 45 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ Q-8750 18.755 ms 1 - 255 MB NPU
PidNet TFLITE float Qualcomm® SA7255P 116.663 ms 0 - 228 MB NPU
PidNet TFLITE float Qualcomm® SA8295P 51.063 ms 3 - 245 MB NPU
PidNet TFLITE float Snapdragon® 8 Elite Mobile 18.755 ms 1 - 255 MB NPU
PidNet TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 12.872 ms 2 - 273 MB NPU
PidNet TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 38.047 ms 1 - 264 MB NPU
PidNet TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 57.98 ms 1 - 265 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 205.616 ms 3 - 72 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 49.771 ms 1 - 18 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 98.37 ms 1 - 214 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 50.103 ms 1 - 3 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA8775P 51.042 ms 1 - 214 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA8650P 51.042 ms 1 - 214 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA8255P 51.042 ms 1 - 214 MB NPU
PidNet TFLITE w8a8 Qualcomm® QCS8450 57.98 ms 1 - 265 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 51.643 ms 1 - 18 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 232.253 ms 2 - 234 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 66.599 ms 3 - 221 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 69.594 ms 1 - 234 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA7255P 98.37 ms 1 - 214 MB NPU
PidNet TFLITE w8a8 Snapdragon® 8 Elite Mobile 69.594 ms 1 - 234 MB NPU
PidNet TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 22.172 ms 1 - 268 MB NPU
PidNet TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 66.599 ms 3 - 221 MB NPU

License

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

References

Community

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Image Segmentation
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Paper for qualcomm/PidNet