Hugging Face's logo

qualcomm
/
FFNet-78S

Image Segmentation
PyTorch
android
Model card Files
xet
 Community

FFNet-78S: Optimized for Qualcomm Devices

FFNet-78S 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-78S 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 FFNet-78S 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-78S on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

  • Input resolution: 2048x1024
  • Model checkpoint: ffnet78S_dBBB_cityscapes_state_dict_quarts
  • Model size (float): 105 MB
  • Model size (w8a8): 26.7 MB
  • Number of output classes: 19
  • Number of parameters: 27.5M

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
FFNet-78S ONNX float Snapdragon® X2 Elite 18.344 ms 24 - 24 MB NPU
FFNet-78S ONNX float Snapdragon® X Elite 37.623 ms 31 - 31 MB NPU
FFNet-78S ONNX float Snapdragon® 8 Gen 3 Mobile 27.156 ms 13 - 331 MB NPU
FFNet-78S ONNX float Snapdragon® 8 Gen 1 Mobile 59.339 ms 1 - 281 MB NPU
FFNet-78S ONNX float Qualcomm® Dragonwing™ IQ-8275 57.981 ms 24 - 51 MB NPU
FFNet-78S ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 38.607 ms 24 - 54 MB NPU
FFNet-78S ONNX float Qualcomm® QCS8450 59.339 ms 1 - 281 MB NPU
FFNet-78S ONNX float Qualcomm® Dragonwing™ IQ-9075 56.6 ms 24 - 51 MB NPU
FFNet-78S ONNX float Qualcomm® Dragonwing™ IQ-X7181 37.623 ms 31 - 31 MB NPU
FFNet-78S ONNX float Qualcomm® Dragonwing™ Q-8750 21.551 ms 6 - 203 MB NPU
FFNet-78S ONNX float Snapdragon® 8 Elite Mobile 21.551 ms 6 - 203 MB NPU
FFNet-78S ONNX float Snapdragon® 8 Elite Gen 5 Mobile 17.756 ms 6 - 231 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® X2 Elite 7.941 ms 7 - 7 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® X Elite 13.482 ms 22 - 22 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 8.539 ms 7 - 298 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 17.807 ms 8 - 300 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 57.248 ms 6 - 9 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® Dragonwing™ IQ-8275 12.772 ms 6 - 10 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 12.765 ms 0 - 244 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCS8450 17.807 ms 8 - 300 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 13.194 ms 6 - 9 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 13.482 ms 22 - 22 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 148.674 ms 5 - 236 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 18.425 ms 7 - 256 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 8.739 ms 1 - 217 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 8 Elite Mobile 8.739 ms 1 - 217 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 7.422 ms 2 - 218 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 18.425 ms 7 - 256 MB NPU
FFNet-78S QNN_DLC float Snapdragon® X2 Elite 18.044 ms 24 - 24 MB NPU
FFNet-78S QNN_DLC float Snapdragon® X Elite 44.202 ms 24 - 24 MB NPU
FFNet-78S QNN_DLC float Snapdragon® 8 Gen 3 Mobile 29.665 ms 19 - 319 MB NPU
FFNet-78S QNN_DLC float Snapdragon® 8 Gen 1 Mobile 85.825 ms 2 - 284 MB NPU
FFNet-78S QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 62.871 ms 24 - 53 MB NPU
FFNet-78S QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 42.689 ms 24 - 26 MB NPU
FFNet-78S QNN_DLC float Qualcomm® SA8775P 60.91 ms 24 - 235 MB NPU
FFNet-78S QNN_DLC float Qualcomm® SA8650P 60.91 ms 24 - 235 MB NPU
FFNet-78S QNN_DLC float Qualcomm® SA8255P 60.91 ms 24 - 235 MB NPU
FFNet-78S QNN_DLC float Qualcomm® QCS8450 85.825 ms 2 - 284 MB NPU
FFNet-78S QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 62.353 ms 26 - 54 MB NPU
FFNet-78S QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 44.202 ms 24 - 24 MB NPU
FFNet-78S QNN_DLC float Qualcomm® Dragonwing™ Q-8750 21.746 ms 0 - 227 MB NPU
FFNet-78S QNN_DLC float Qualcomm® SA7255P 187.925 ms 24 - 234 MB NPU
FFNet-78S QNN_DLC float Qualcomm® SA8295P 66.501 ms 24 - 227 MB NPU
FFNet-78S QNN_DLC float Snapdragon® 8 Elite Mobile 21.746 ms 0 - 227 MB NPU
FFNet-78S QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 15.697 ms 14 - 266 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® X2 Elite 7.017 ms 6 - 6 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® X Elite 17.778 ms 6 - 6 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 11.634 ms 6 - 292 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 24.045 ms 6 - 292 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS6490 63.743 ms 4 - 12 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 16.491 ms 4 - 13 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 16.784 ms 6 - 8 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® SA8775P 17.164 ms 6 - 226 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® SA8650P 17.164 ms 6 - 226 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® SA8255P 17.164 ms 6 - 226 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS8450 24.045 ms 6 - 292 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 17.495 ms 6 - 14 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 17.778 ms 6 - 6 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-6690 166.587 ms 6 - 254 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-7790 22.254 ms 6 - 243 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 8.068 ms 6 - 239 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® SA7255P 39.078 ms 6 - 225 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® SA8295P 23.327 ms 6 - 227 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 8.068 ms 6 - 239 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 5.962 ms 6 - 252 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 22.254 ms 6 - 243 MB NPU
FFNet-78S TFLITE float Snapdragon® 8 Gen 3 Mobile 29.198 ms 2 - 378 MB NPU
FFNet-78S TFLITE float Snapdragon® 8 Gen 1 Mobile 86.055 ms 1 - 361 MB NPU
FFNet-78S TFLITE float Qualcomm® Dragonwing™ IQ-8275 62.61 ms 0 - 83 MB NPU
FFNet-78S TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 42.7 ms 3 - 5 MB NPU
FFNet-78S TFLITE float Qualcomm® SA8775P 61.005 ms 3 - 246 MB NPU
FFNet-78S TFLITE float Qualcomm® SA8650P 61.005 ms 3 - 246 MB NPU
FFNet-78S TFLITE float Qualcomm® SA8255P 61.005 ms 3 - 246 MB NPU
FFNet-78S TFLITE float Qualcomm® QCS8450 86.055 ms 1 - 361 MB NPU
FFNet-78S TFLITE float Qualcomm® Dragonwing™ IQ-9075 62.094 ms 0 - 82 MB NPU
FFNet-78S TFLITE float Qualcomm® Dragonwing™ Q-8750 21.747 ms 1 - 262 MB NPU
FFNet-78S TFLITE float Qualcomm® SA7255P 188.025 ms 3 - 245 MB NPU
FFNet-78S TFLITE float Qualcomm® SA8295P 66.497 ms 3 - 240 MB NPU
FFNet-78S TFLITE float Snapdragon® 8 Elite Mobile 21.747 ms 1 - 262 MB NPU
FFNet-78S TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 15.79 ms 0 - 280 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 6.444 ms 0 - 286 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 18.486 ms 4 - 292 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 49.851 ms 1 - 37 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 9.775 ms 1 - 37 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 9.094 ms 1 - 3 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® SA8775P 9.756 ms 1 - 219 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® SA8650P 9.756 ms 1 - 219 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® SA8255P 9.756 ms 1 - 219 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® QCS8450 18.486 ms 4 - 292 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 9.498 ms 0 - 36 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 138.43 ms 1 - 246 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 14.674 ms 1 - 234 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 5.032 ms 0 - 230 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® SA7255P 26.255 ms 1 - 218 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® SA8295P 14.88 ms 0 - 221 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 8 Elite Mobile 5.032 ms 0 - 230 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 3.658 ms 1 - 243 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 14.674 ms 1 - 234 MB NPU

License

  • The license for the original implementation of FFNet-78S 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-78S