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 2.50, 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-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:

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

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
FFNet-40S ONNX float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 13.181 ms 7 - 208 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Elite For Galaxy Mobile 17.142 ms 6 - 183 MB NPU
FFNet-40S ONNX float Snapdragon® X2 Elite 14.047 ms 24 - 24 MB NPU
FFNet-40S ONNX float Snapdragon® X Elite 30.615 ms 24 - 24 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Gen 3 Mobile 25.141 ms 29 - 285 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Gen 1 Mobile 46.346 ms 29 - 280 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ IQ-8275 43.501 ms 24 - 28 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 31.845 ms 24 - 54 MB NPU
FFNet-40S ONNX float Qualcomm® QCS8450 46.346 ms 29 - 280 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ IQ-9075 43.461 ms 24 - 27 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ IQ-X7181 30.615 ms 24 - 24 MB NPU
FFNet-40S ONNX float Qualcomm® Dragonwing™ Q-8750 17.142 ms 6 - 183 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 3.566 ms 2 - 190 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite For Galaxy Mobile 4.915 ms 2 - 186 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X2 Elite 3.894 ms 7 - 7 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X Elite 10.604 ms 9 - 9 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 7.14 ms 6 - 243 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 14.033 ms 7 - 245 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 43.292 ms 6 - 9 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ IQ-8275 10.404 ms 6 - 10 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 10.042 ms 0 - 163 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS8450 14.033 ms 7 - 245 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 11.641 ms 6 - 9 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 10.604 ms 9 - 9 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 108.865 ms 5 - 199 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 14.465 ms 7 - 206 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 4.915 ms 2 - 186 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 14.465 ms 7 - 206 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 12.405 ms 18 - 236 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite For Galaxy Mobile 18.011 ms 0 - 203 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X2 Elite 14.763 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X Elite 38.202 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Gen 3 Mobile 26.224 ms 24 - 288 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Gen 1 Mobile 50.051 ms 24 - 281 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ IQ-8275 48.92 ms 24 - 53 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 36.423 ms 24 - 26 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8775P 49.196 ms 24 - 210 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8650P 49.196 ms 24 - 210 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8255P 49.196 ms 24 - 210 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8450 50.051 ms 24 - 281 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 50.27 ms 24 - 52 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 38.202 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Qualcomm® Dragonwing™ Q-8750 18.011 ms 0 - 203 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA7255P 136.882 ms 24 - 209 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8295P 53.75 ms 24 - 212 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 5.398 ms 6 - 222 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite For Galaxy Mobile 7.186 ms 6 - 206 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X2 Elite 6.277 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X Elite 16.147 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 10.503 ms 6 - 244 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 17.84 ms 6 - 243 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS6490 49.449 ms 4 - 13 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-8275 14.984 ms 6 - 15 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 15.153 ms 6 - 8 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8450 17.84 ms 6 - 243 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 16.173 ms 6 - 15 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 16.147 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-6690 134.32 ms 6 - 194 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-7790 18.771 ms 6 - 206 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 7.186 ms 6 - 206 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA7255P 33.399 ms 6 - 193 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8295P 20.14 ms 6 - 196 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 18.771 ms 6 - 206 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 12.462 ms 2 - 231 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite For Galaxy Mobile 18.135 ms 1 - 216 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Gen 3 Mobile 26.163 ms 2 - 298 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Gen 1 Mobile 49.977 ms 2 - 294 MB NPU
FFNet-40S TFLITE float Qualcomm® Dragonwing™ IQ-8275 48.889 ms 0 - 57 MB NPU
FFNet-40S TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 36.428 ms 0 - 167 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS8450 49.977 ms 2 - 294 MB NPU
FFNet-40S TFLITE float Qualcomm® Dragonwing™ IQ-9075 50.295 ms 0 - 56 MB NPU
FFNet-40S TFLITE float Qualcomm® Dragonwing™ Q-8750 18.135 ms 1 - 216 MB NPU
FFNet-40S TFLITE float Qualcomm® SA7255P 136.794 ms 3 - 202 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8295P 53.908 ms 0 - 201 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 3.05 ms 0 - 215 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite For Galaxy Mobile 3.986 ms 0 - 202 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 5.323 ms 1 - 237 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 9.524 ms 1 - 240 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 33.319 ms 0 - 23 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ IQ-8275 8.103 ms 1 - 24 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 7.518 ms 1 - 3 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8775P 8.273 ms 1 - 188 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8650P 8.273 ms 1 - 188 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8255P 8.273 ms 1 - 188 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8450 9.524 ms 1 - 240 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 8.074 ms 1 - 24 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 106.125 ms 0 - 186 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 11.346 ms 0 - 202 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 3.986 ms 0 - 202 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA7255P 20.173 ms 1 - 187 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8295P 11.702 ms 0 - 190 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 11.346 ms 0 - 202 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