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 1.25.0 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-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:

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

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
FFNet-78S ONNX float Snapdragon® X2 Elite 18.424 ms 188 - 188 MB NPU
FFNet-78S ONNX float Snapdragon® X Elite 37.649 ms 158 - 158 MB NPU
FFNet-78S ONNX float Snapdragon® 8 Gen 3 Mobile 27.006 ms 37 - 340 MB NPU
FFNet-78S ONNX float Snapdragon® 8 Gen 1 Mobile 59.831 ms 29 - 299 MB NPU
FFNet-78S ONNX float Qualcomm® QCS8550 (Proxy) 38.38 ms 24 - 56 MB NPU
FFNet-78S ONNX float Qualcomm® QCS8450 59.831 ms 29 - 299 MB NPU
FFNet-78S ONNX float Snapdragon® 8 Elite Mobile 21.241 ms 5 - 203 MB NPU
FFNet-78S ONNX float Snapdragon® 8 Elite Gen 5 Mobile 17.628 ms 5 - 223 MB NPU
FFNet-78S ONNX float Qualcomm® QCS9075 59.348 ms 24 - 93 MB NPU
FFNet-78S ONNX float Qualcomm® QCS8750 21.241 ms 5 - 203 MB NPU
FFNet-78S ONNX float Qualcomm® QCS7181 37.649 ms 158 - 158 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® X2 Elite 7.933 ms 207 - 207 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® X Elite 13.392 ms 175 - 175 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 8.533 ms 7 - 291 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 17.79 ms 7 - 292 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCS6490 65.773 ms 6 - 51 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCS8550 (Proxy) 12.897 ms 6 - 10 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCS8450 17.79 ms 7 - 292 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 7.374 ms 2 - 216 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 17.967 ms 7 - 252 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCM6690 148.392 ms 5 - 233 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCS9075 14.964 ms 6 - 51 MB NPU
FFNet-78S ONNX w8a8 Snapdragon® 8 Elite Mobile 8.75 ms 1 - 215 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCS7790 17.967 ms 7 - 252 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCS8750 8.75 ms 1 - 215 MB NPU
FFNet-78S ONNX w8a8 Qualcomm® QCS7181 13.392 ms 175 - 175 MB NPU
FFNet-78S QNN_DLC float Snapdragon® X2 Elite 18.08 ms 24 - 24 MB NPU
FFNet-78S QNN_DLC float Snapdragon® X Elite 44.301 ms 24 - 24 MB NPU
FFNet-78S QNN_DLC float Snapdragon® 8 Gen 3 Mobile 29.229 ms 23 - 319 MB NPU
FFNet-78S QNN_DLC float Snapdragon® 8 Gen 1 Mobile 86.142 ms 25 - 308 MB NPU
FFNet-78S QNN_DLC float Qualcomm® QCS8275 188.046 ms 24 - 234 MB NPU
FFNet-78S QNN_DLC float Qualcomm® QCS8550 (Proxy) 42.703 ms 24 - 54 MB NPU
FFNet-78S QNN_DLC float Qualcomm® QCS8450 86.142 ms 25 - 308 MB NPU
FFNet-78S QNN_DLC float Snapdragon® 8 Elite Mobile 21.675 ms 12 - 239 MB NPU
FFNet-78S QNN_DLC float Qualcomm® SA8295P 66.472 ms 24 - 229 MB NPU
FFNet-78S QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 15.464 ms 0 - 253 MB NPU
FFNet-78S QNN_DLC float Qualcomm® SA7255P 188.046 ms 24 - 234 MB NPU
FFNet-78S QNN_DLC float Qualcomm® QCS9075 73.292 ms 24 - 52 MB NPU
FFNet-78S QNN_DLC float Qualcomm® QCS8750 21.675 ms 12 - 239 MB NPU
FFNet-78S QNN_DLC float Qualcomm® QCS7181 44.301 ms 24 - 24 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® X2 Elite 6.961 ms 6 - 6 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® X Elite 17.79 ms 6 - 6 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 11.605 ms 6 - 289 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 23.627 ms 6 - 294 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS6490 72.672 ms 6 - 14 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS8275 39.031 ms 6 - 226 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS8550 (Proxy) 16.857 ms 6 - 179 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS8450 23.627 ms 6 - 294 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 5.914 ms 6 - 252 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 22.186 ms 6 - 243 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCM6690 166.265 ms 6 - 255 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS9075 19.99 ms 6 - 14 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® SA7255P 39.031 ms 6 - 226 MB NPU
FFNet-78S QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 8.064 ms 6 - 240 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® SA8295P 23.341 ms 6 - 227 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS7790 22.186 ms 6 - 243 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS8750 8.064 ms 6 - 240 MB NPU
FFNet-78S QNN_DLC w8a8 Qualcomm® QCS7181 17.79 ms 6 - 6 MB NPU
FFNet-78S TFLITE float Snapdragon® 8 Gen 3 Mobile 29.494 ms 2 - 377 MB NPU
FFNet-78S TFLITE float Snapdragon® 8 Gen 1 Mobile 86.383 ms 3 - 362 MB NPU
FFNet-78S TFLITE float Qualcomm® QCS8275 187.959 ms 3 - 244 MB NPU
FFNet-78S TFLITE float Qualcomm® QCS8550 (Proxy) 42.356 ms 2 - 5 MB NPU
FFNet-78S TFLITE float Qualcomm® SA8775P 355.8 ms 3 - 32 MB GPU
FFNet-78S TFLITE float Qualcomm® SA8650P 355.8 ms 3 - 32 MB GPU
FFNet-78S TFLITE float Qualcomm® SA8255P 355.8 ms 3 - 32 MB GPU
FFNet-78S TFLITE float Qualcomm® QCS8450 86.383 ms 3 - 362 MB NPU
FFNet-78S TFLITE float Snapdragon® 8 Elite Mobile 21.617 ms 2 - 264 MB NPU
FFNet-78S TFLITE float Qualcomm® SA8295P 66.46 ms 3 - 240 MB NPU
FFNet-78S TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 15.48 ms 2 - 284 MB NPU
FFNet-78S TFLITE float Qualcomm® SA7255P 187.959 ms 3 - 244 MB NPU
FFNet-78S TFLITE float Qualcomm® QCS9075 72.893 ms 0 - 82 MB NPU
FFNet-78S TFLITE float Qualcomm® QCS8750 21.617 ms 2 - 264 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 6.43 ms 0 - 287 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 18.183 ms 1 - 289 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® QCS6490 57.256 ms 1 - 36 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® QCS8275 26.238 ms 1 - 218 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® QCS8550 (Proxy) 9.077 ms 1 - 3 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® SA8775P 648.725 ms 26 - 56 MB GPU
FFNet-78S TFLITE w8a8 Qualcomm® SA8650P 648.725 ms 26 - 56 MB GPU
FFNet-78S TFLITE w8a8 Qualcomm® SA8255P 648.725 ms 26 - 56 MB GPU
FFNet-78S TFLITE w8a8 Qualcomm® QCS8450 18.183 ms 1 - 289 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 3.616 ms 1 - 245 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 14.513 ms 1 - 234 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® QCM6690 146.356 ms 1 - 246 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® QCS9075 11.16 ms 1 - 36 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® SA7255P 26.238 ms 1 - 218 MB NPU
FFNet-78S TFLITE w8a8 Snapdragon® 8 Elite Mobile 5.041 ms 0 - 230 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® SA8295P 14.896 ms 1 - 222 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® QCS7790 14.513 ms 1 - 234 MB NPU
FFNet-78S TFLITE w8a8 Qualcomm® QCS8750 5.041 ms 0 - 230 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