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 2.50, ONNX Runtime 1.30.0 Download
ONNX w8a8 Universal QAIRT 2.50, ONNX Runtime 1.30.0 Download
QNN_DLC float Universal QAIRT 2.50 Download
QNN_DLC w8a8 Universal QAIRT 2.50 Download
TFLITE float Universal QAIRT 2.50 Download
TFLITE w8a8 Universal QAIRT 2.50 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® 8 Elite Gen 5 For Galaxy Mobile 16.269 ms 7 - 245 MB NPU
PidNet ONNX float Snapdragon® 8 Elite For Galaxy Mobile 17.692 ms 6 - 212 MB NPU
PidNet ONNX float Snapdragon® X2 Elite 13.333 ms 24 - 24 MB NPU
PidNet ONNX float Snapdragon® X Elite 31.477 ms 24 - 24 MB NPU
PidNet ONNX float Snapdragon® 8 Gen 3 Mobile 22.111 ms 31 - 325 MB NPU
PidNet ONNX float Snapdragon® 8 Gen 1 Mobile 47.204 ms 26 - 318 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ IQ-8275 39.709 ms 24 - 52 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 32.464 ms 24 - 26 MB NPU
PidNet ONNX float Qualcomm® QCS8450 47.204 ms 26 - 318 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ IQ-9075 41.249 ms 24 - 51 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ IQ-X7181 31.477 ms 24 - 24 MB NPU
PidNet ONNX float Qualcomm® Dragonwing™ Q-8750 17.692 ms 6 - 212 MB NPU
PidNet ONNX w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 22.941 ms 3 - 230 MB NPU
PidNet ONNX w8a8 Snapdragon® 8 Elite For Galaxy Mobile 41.389 ms 2 - 211 MB NPU
PidNet ONNX w8a8 Snapdragon® X2 Elite 23.206 ms 7 - 7 MB NPU
PidNet ONNX w8a8 Snapdragon® X Elite 54.847 ms 7 - 7 MB NPU
PidNet ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 39.078 ms 7 - 263 MB NPU
PidNet ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 55.241 ms 7 - 262 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ IQ-8275 50.701 ms 6 - 10 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 52.403 ms 5 - 10 MB NPU
PidNet ONNX w8a8 Qualcomm® QCS8450 55.241 ms 7 - 262 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 54.11 ms 6 - 9 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 54.847 ms 7 - 7 MB NPU
PidNet ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 41.389 ms 2 - 211 MB NPU
PidNet QNN_DLC float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 12.51 ms 22 - 274 MB NPU
PidNet QNN_DLC float Snapdragon® 8 Elite For Galaxy Mobile 18.52 ms 18 - 250 MB NPU
PidNet QNN_DLC float Snapdragon® X2 Elite 14.04 ms 24 - 24 MB NPU
PidNet QNN_DLC float Snapdragon® X Elite 38.094 ms 24 - 24 MB NPU
PidNet QNN_DLC float Snapdragon® 8 Gen 3 Mobile 25.228 ms 18 - 313 MB NPU
PidNet QNN_DLC float Snapdragon® 8 Gen 1 Mobile 50.989 ms 0 - 293 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 44.075 ms 24 - 53 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 36.818 ms 24 - 26 MB NPU
PidNet QNN_DLC float Qualcomm® SA8775P 45.803 ms 24 - 232 MB NPU
PidNet QNN_DLC float Qualcomm® SA8650P 45.803 ms 24 - 232 MB NPU
PidNet QNN_DLC float Qualcomm® SA8255P 45.803 ms 24 - 232 MB NPU
PidNet QNN_DLC float Qualcomm® QCS8450 50.989 ms 0 - 293 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 47.131 ms 24 - 52 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 38.094 ms 24 - 24 MB NPU
PidNet QNN_DLC float Qualcomm® Dragonwing™ Q-8750 18.52 ms 18 - 250 MB NPU
PidNet QNN_DLC float Qualcomm® SA8295P 52.748 ms 24 - 242 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 24.517 ms 6 - 256 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® 8 Elite For Galaxy Mobile 44.552 ms 6 - 229 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® X2 Elite 25.101 ms 6 - 6 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® X Elite 60.442 ms 6 - 6 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 43.06 ms 6 - 255 MB NPU
PidNet QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 59.27 ms 6 - 257 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 56.334 ms 3 - 12 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 57.512 ms 6 - 8 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA8775P 58.537 ms 6 - 208 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA8650P 58.537 ms 6 - 208 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA8255P 58.537 ms 6 - 208 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® QCS8450 59.27 ms 6 - 257 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 59.263 ms 6 - 15 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 60.442 ms 6 - 6 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 44.552 ms 6 - 229 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA7255P 110.669 ms 6 - 207 MB NPU
PidNet QNN_DLC w8a8 Qualcomm® SA8295P 66.505 ms 6 - 212 MB NPU
PidNet TFLITE float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 12.7 ms 2 - 257 MB NPU
PidNet TFLITE float Snapdragon® 8 Elite For Galaxy Mobile 18.304 ms 1 - 231 MB NPU
PidNet TFLITE float Snapdragon® 8 Gen 3 Mobile 25.623 ms 2 - 310 MB NPU
PidNet TFLITE float Snapdragon® 8 Gen 1 Mobile 49.312 ms 0 - 311 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ IQ-8275 43.369 ms 0 - 46 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 35.846 ms 3 - 5 MB NPU
PidNet TFLITE float Qualcomm® SA8775P 45.157 ms 0 - 215 MB NPU
PidNet TFLITE float Qualcomm® SA8650P 45.157 ms 0 - 215 MB NPU
PidNet TFLITE float Qualcomm® SA8255P 45.157 ms 0 - 215 MB NPU
PidNet TFLITE float Qualcomm® QCS8450 49.312 ms 0 - 311 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ IQ-9075 46.034 ms 0 - 45 MB NPU
PidNet TFLITE float Qualcomm® Dragonwing™ Q-8750 18.304 ms 1 - 231 MB NPU
PidNet TFLITE float Qualcomm® SA7255P 116.121 ms 3 - 218 MB NPU
PidNet TFLITE float Qualcomm® SA8295P 52.267 ms 2 - 232 MB NPU
PidNet TFLITE w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 22.155 ms 1 - 249 MB NPU
PidNet TFLITE w8a8 Snapdragon® 8 Elite For Galaxy Mobile 69.026 ms 1 - 224 MB NPU
PidNet TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 37.818 ms 0 - 251 MB NPU
PidNet TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 51.075 ms 1 - 256 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 163.044 ms 1 - 71 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 49.666 ms 1 - 19 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 50.152 ms 1 - 4 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA8775P 50.98 ms 1 - 204 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA8650P 50.98 ms 1 - 204 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA8255P 50.98 ms 1 - 204 MB NPU
PidNet TFLITE w8a8 Qualcomm® QCS8450 51.075 ms 1 - 256 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 51.498 ms 0 - 17 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 239.105 ms 2 - 187 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 66.261 ms 2 - 213 MB NPU
PidNet TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 69.026 ms 1 - 224 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA7255P 98.069 ms 1 - 203 MB NPU
PidNet TFLITE w8a8 Qualcomm® SA8295P 57.86 ms 1 - 207 MB NPU
PidNet TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 66.261 ms 2 - 213 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