Hugging Face's logo

qualcomm
/
FFNet-78S-LowRes

Image Segmentation
PyTorch
real_time
android
Model card Files
xet
 Community

FFNet-78S-LowRes: Optimized for Qualcomm Devices

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

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

  • Input resolution: 1024x512
  • Model checkpoint: ffnet78S_BCC_cityscapes_state_dict_quarts_pre_down
  • Model size (float): 102 MB
  • Model size (w8a8): 26.0 MB
  • Number of output classes: 19
  • Number of parameters: 26.8M

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
FFNet-78S-LowRes ONNX float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 8.298 ms 8 - 186 MB NPU
FFNet-78S-LowRes ONNX float Snapdragon® 8 Elite For Galaxy Mobile 10.662 ms 7 - 181 MB NPU
FFNet-78S-LowRes ONNX float Snapdragon® X2 Elite 10.046 ms 10 - 10 MB NPU
FFNet-78S-LowRes ONNX float Snapdragon® X Elite 19.224 ms 47 - 47 MB NPU
FFNet-78S-LowRes ONNX float Snapdragon® 8 Gen 3 Mobile 13.497 ms 13 - 249 MB NPU
FFNet-78S-LowRes ONNX float Snapdragon® 8 Gen 1 Mobile 33.298 ms 11 - 239 MB NPU
FFNet-78S-LowRes ONNX float Qualcomm® Dragonwing™ IQ-8275 37.802 ms 6 - 16 MB NPU
FFNet-78S-LowRes ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 18.479 ms 0 - 286 MB NPU
FFNet-78S-LowRes ONNX float Qualcomm® QCS8450 33.298 ms 11 - 239 MB NPU
FFNet-78S-LowRes ONNX float Qualcomm® Dragonwing™ IQ-9075 33.274 ms 6 - 15 MB NPU
FFNet-78S-LowRes ONNX float Qualcomm® Dragonwing™ IQ-X7181 19.224 ms 47 - 47 MB NPU
FFNet-78S-LowRes ONNX float Qualcomm® Dragonwing™ Q-8750 10.662 ms 7 - 181 MB NPU
FFNet-78S-LowRes ONNX w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 2.267 ms 0 - 198 MB NPU
FFNet-78S-LowRes ONNX w8a8 Snapdragon® 8 Elite For Galaxy Mobile 3.044 ms 0 - 191 MB NPU
FFNet-78S-LowRes ONNX w8a8 Snapdragon® X2 Elite 2.7 ms 4 - 4 MB NPU
FFNet-78S-LowRes ONNX w8a8 Snapdragon® X Elite 5.718 ms 25 - 25 MB NPU
FFNet-78S-LowRes ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 3.858 ms 3 - 248 MB NPU
FFNet-78S-LowRes ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 6.901 ms 3 - 249 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 24.979 ms 2 - 5 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® Dragonwing™ IQ-8275 6.056 ms 2 - 6 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 5.45 ms 2 - 5 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® QCS8450 6.901 ms 3 - 249 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 5.519 ms 2 - 5 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 5.718 ms 25 - 25 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 100.739 ms 0 - 202 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 12.193 ms 4 - 212 MB NPU
FFNet-78S-LowRes ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 3.044 ms 0 - 191 MB NPU
FFNet-78S-LowRes ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 12.193 ms 4 - 212 MB NPU
FFNet-78S-LowRes QNN_DLC float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 10.959 ms 6 - 194 MB NPU
FFNet-78S-LowRes QNN_DLC float Snapdragon® 8 Elite For Galaxy Mobile 11.337 ms 0 - 180 MB NPU
FFNet-78S-LowRes QNN_DLC float Snapdragon® X2 Elite 12.099 ms 6 - 6 MB NPU
FFNet-78S-LowRes QNN_DLC float Snapdragon® X Elite 22.029 ms 6 - 6 MB NPU
FFNet-78S-LowRes QNN_DLC float Snapdragon® 8 Gen 3 Mobile 17.935 ms 6 - 238 MB NPU
FFNet-78S-LowRes QNN_DLC float Snapdragon® 8 Gen 1 Mobile 34.975 ms 6 - 227 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 40.127 ms 6 - 17 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 21.039 ms 6 - 8 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® SA8775P 35.797 ms 1 - 174 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® SA8650P 35.797 ms 1 - 174 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® SA8255P 35.797 ms 1 - 174 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® QCS8450 34.975 ms 6 - 227 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 35.718 ms 6 - 16 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 22.029 ms 6 - 6 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® Dragonwing™ Q-8750 11.337 ms 0 - 180 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® SA7255P 131.786 ms 0 - 174 MB NPU
FFNet-78S-LowRes QNN_DLC float Qualcomm® SA8295P 38.115 ms 0 - 171 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 2.662 ms 2 - 190 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Snapdragon® 8 Elite For Galaxy Mobile 3.54 ms 2 - 185 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Snapdragon® X2 Elite 3.455 ms 2 - 2 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Snapdragon® X Elite 7.67 ms 2 - 2 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 4.934 ms 2 - 130 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 8.424 ms 2 - 131 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS6490 26.179 ms 4 - 8 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 7.447 ms 2 - 7 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 6.996 ms 2 - 3 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® SA8775P 7.357 ms 2 - 82 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® SA8650P 7.357 ms 2 - 82 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® SA8255P 7.357 ms 2 - 82 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® QCS8450 8.424 ms 2 - 131 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 7.174 ms 3 - 7 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 7.67 ms 2 - 2 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-6690 108.697 ms 2 - 196 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-7790 13.362 ms 2 - 209 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 3.54 ms 2 - 185 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® SA7255P 20.424 ms 2 - 79 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Qualcomm® SA8295P 11.419 ms 0 - 76 MB NPU
FFNet-78S-LowRes QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 13.362 ms 2 - 209 MB NPU
FFNet-78S-LowRes TFLITE float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 10.984 ms 2 - 222 MB NPU
FFNet-78S-LowRes TFLITE float Snapdragon® 8 Elite For Galaxy Mobile 11.353 ms 2 - 214 MB NPU
FFNet-78S-LowRes TFLITE float Snapdragon® 8 Gen 3 Mobile 17.913 ms 1 - 314 MB NPU
FFNet-78S-LowRes TFLITE float Snapdragon® 8 Gen 1 Mobile 34.949 ms 2 - 301 MB NPU
FFNet-78S-LowRes TFLITE float Qualcomm® Dragonwing™ IQ-8275 40.314 ms 0 - 64 MB NPU
FFNet-78S-LowRes TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 21.279 ms 2 - 5 MB NPU
FFNet-78S-LowRes TFLITE float Qualcomm® QCS8450 34.949 ms 2 - 301 MB NPU
FFNet-78S-LowRes TFLITE float Qualcomm® Dragonwing™ IQ-9075 36.332 ms 0 - 63 MB NPU
FFNet-78S-LowRes TFLITE float Qualcomm® Dragonwing™ Q-8750 11.353 ms 2 - 214 MB NPU
FFNet-78S-LowRes TFLITE float Qualcomm® SA7255P 131.88 ms 3 - 210 MB NPU
FFNet-78S-LowRes TFLITE float Qualcomm® SA8295P 38.113 ms 2 - 207 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 2.089 ms 0 - 186 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Snapdragon® 8 Elite For Galaxy Mobile 2.77 ms 0 - 184 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 3.661 ms 8 - 141 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 6.529 ms 0 - 133 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 22.299 ms 1 - 31 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 5.83 ms 1 - 32 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 5.199 ms 1 - 3 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® SA8775P 5.51 ms 1 - 79 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® SA8650P 5.51 ms 1 - 79 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® SA8255P 5.51 ms 1 - 79 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® QCS8450 6.529 ms 0 - 133 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 5.303 ms 1 - 31 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 102.529 ms 1 - 194 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 11.645 ms 1 - 201 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 2.77 ms 0 - 184 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® SA7255P 17.272 ms 1 - 78 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Qualcomm® SA8295P 9.455 ms 1 - 75 MB NPU
FFNet-78S-LowRes TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 11.645 ms 1 - 201 MB NPU

License

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