Hugging Face's logo

qualcomm
/
FFNet-40S

Image Segmentation
PyTorch
real_time
android
Model card Files
xet
 Community

FFNet-40S: Optimized for Qualcomm Devices

FFNet-40S is a "fuss-free network" that segments street scene images with per-pixel classes like road, sidewalk, and pedestrian. Trained on the Cityscapes dataset.

This is based on the implementation of FFNet-40S 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 1.25.0 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 FFNet-40S 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 FFNet-40S on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

  • Model checkpoint: ffnet40S_dBBB_cityscapes_state_dict_quarts
  • Input resolution: 2048x1024
  • Number of output classes: 19
  • Number of parameters: 13.9M
  • Model size (float): 53.1 MB
  • Model size (w8a8): 13.5 MB

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
FFNet-40S ONNX float Snapdragon® X2 Elite 14.119 ms 24 - 24 MB NPU
FFNet-40S ONNX float Snapdragon® X Elite 31.29 ms 24 - 24 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Gen 3 Mobile 22.344 ms 30 - 316 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Gen 1 Mobile 52.093 ms 24 - 282 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 32.039 ms 0 - 30 MB NPU
FFNet-40S ONNX float Qualcomm® QCS8450 52.093 ms 24 - 282 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ IQ-9075 44.299 ms 24 - 27 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Elite Gen 5 Mobile 13.485 ms 5 - 213 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Elite Mobile 17.737 ms 8 - 192 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ Q-8750 17.737 ms 8 - 192 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ IQ-X7181 31.29 ms 24 - 24 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X2 Elite 6.922 ms 7 - 7 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X Elite 11.714 ms 8 - 8 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 7.386 ms 7 - 253 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 15.783 ms 7 - 255 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 49.048 ms 6 - 9 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 11.195 ms 6 - 44 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS8450 15.783 ms 7 - 255 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite Mobile 7.686 ms 1 - 193 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 14.779 ms 7 - 223 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 6.619 ms 1 - 194 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 15.541 ms 6 - 9 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 110.533 ms 5 - 242 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 14.779 ms 7 - 223 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 7.686 ms 1 - 193 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 11.714 ms 8 - 8 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X2 Elite 14.568 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X Elite 38.033 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Gen 3 Mobile 24.996 ms 23 - 294 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Gen 1 Mobile 63.509 ms 24 - 290 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8275 135.546 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 36.634 ms 24 - 238 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8775P 48.951 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8650P 48.951 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8255P 48.951 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8450 63.509 ms 24 - 290 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 49.857 ms 24 - 52 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 12.712 ms 16 - 257 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA7255P 135.546 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite Mobile 18.075 ms 8 - 231 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8295P 53.704 ms 24 - 224 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ Q-8750 18.075 ms 8 - 231 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 38.033 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X2 Elite 6.076 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X Elite 16.016 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 10.514 ms 6 - 252 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 21.266 ms 6 - 254 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS6490 55.583 ms 3 - 11 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8275 32.894 ms 6 - 202 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 15.042 ms 6 - 9 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8775P 15.673 ms 6 - 203 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8650P 15.673 ms 6 - 203 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8255P 15.673 ms 6 - 203 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8450 21.266 ms 6 - 254 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 7.064 ms 6 - 217 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 18.839 ms 6 - 225 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 5.295 ms 6 - 233 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 18.598 ms 8 - 16 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-6690 126.294 ms 6 - 238 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA7255P 32.894 ms 6 - 202 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8295P 20.04 ms 6 - 206 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-7790 18.839 ms 6 - 225 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 7.064 ms 6 - 217 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 16.016 ms 6 - 6 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Gen 3 Mobile 25.134 ms 2 - 306 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Gen 1 Mobile 63.419 ms 2 - 303 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS8275 135.654 ms 0 - 207 MB NPU
FFNet-40S TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 36.383 ms 2 - 5 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8775P 48.977 ms 0 - 208 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8650P 48.977 ms 0 - 208 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8255P 48.977 ms 0 - 208 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS8450 63.419 ms 2 - 303 MB NPU
FFNet-40S TFLITE float Qualcomm® Dragonwing™ IQ-9075 49.697 ms 0 - 56 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 12.743 ms 2 - 253 MB NPU
FFNet-40S TFLITE float Qualcomm® SA7255P 135.654 ms 0 - 207 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite Mobile 18.121 ms 0 - 235 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8295P 53.677 ms 3 - 216 MB NPU
FFNet-40S TFLITE float Qualcomm® Dragonwing™ Q-8750 18.121 ms 0 - 235 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 5.455 ms 0 - 247 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 12.37 ms 0 - 247 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 41.174 ms 1 - 23 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8275 20.07 ms 1 - 196 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 7.489 ms 1 - 3 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8775P 8.268 ms 1 - 197 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8650P 8.268 ms 1 - 197 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8255P 8.268 ms 1 - 197 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8450 12.37 ms 0 - 247 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite Mobile 3.991 ms 1 - 209 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 11.468 ms 0 - 216 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 2.973 ms 1 - 226 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 8.019 ms 1 - 23 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 101.456 ms 1 - 232 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA7255P 20.07 ms 1 - 196 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8295P 11.687 ms 1 - 200 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 11.468 ms 0 - 216 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 3.991 ms 1 - 209 MB NPU

License

  • The license for the original implementation of FFNet-40S 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/FFNet-40S