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 14.299 ms 189 - 189 MB NPU
FFNet-40S ONNX float Snapdragon® X Elite 31.56 ms 156 - 156 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Gen 3 Mobile 22.633 ms 30 - 305 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Gen 1 Mobile 51.944 ms 29 - 277 MB NPU
FFNet-40S ONNX float Qualcomm® QCS8550 (Proxy) 31.78 ms 25 - 32 MB NPU
FFNet-40S ONNX float Qualcomm® QCS8450 51.944 ms 29 - 277 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Elite Mobile 17.77 ms 5 - 189 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Elite Gen 5 Mobile 13.353 ms 7 - 211 MB NPU
FFNet-40S ONNX float Qualcomm® QCS9075 47.843 ms 24 - 69 MB NPU
FFNet-40S ONNX float Qualcomm® QCS8750 17.77 ms 5 - 189 MB NPU
FFNet-40S ONNX float Qualcomm® QCS7181 31.56 ms 156 - 156 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X2 Elite 6.905 ms 207 - 207 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X Elite 11.804 ms 176 - 176 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 7.398 ms 7 - 244 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 15.771 ms 7 - 244 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS6490 60.809 ms 6 - 51 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS8550 (Proxy) 11.278 ms 6 - 18 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS8450 15.771 ms 7 - 244 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 6.588 ms 1 - 191 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 14.763 ms 9 - 220 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCM6690 116.409 ms 5 - 239 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS9075 11.973 ms 6 - 51 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite Mobile 7.671 ms 2 - 191 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS7790 14.763 ms 9 - 220 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS8750 7.671 ms 2 - 191 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS7181 11.804 ms 176 - 176 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X2 Elite 14.623 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X Elite 37.818 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Gen 3 Mobile 24.765 ms 24 - 293 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Gen 1 Mobile 63.655 ms 24 - 291 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8275 135.553 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8550 (Proxy) 36.407 ms 24 - 26 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8450 63.655 ms 24 - 291 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite Mobile 18.142 ms 12 - 235 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8295P 53.718 ms 24 - 224 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 12.584 ms 22 - 262 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA7255P 135.553 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS9075 62.171 ms 24 - 52 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8750 18.142 ms 12 - 235 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS7181 37.818 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X2 Elite 6.147 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X Elite 15.985 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 10.506 ms 6 - 253 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 21.378 ms 6 - 253 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS6490 66.516 ms 6 - 14 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8275 32.926 ms 6 - 202 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8550 (Proxy) 15.053 ms 6 - 8 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8450 21.378 ms 6 - 253 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 5.263 ms 6 - 233 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 18.866 ms 6 - 225 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCM6690 126.521 ms 6 - 240 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS9075 18.598 ms 8 - 16 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA7255P 32.926 ms 6 - 202 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 7.093 ms 6 - 217 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8295P 20.076 ms 6 - 206 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS7790 18.866 ms 6 - 225 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8750 7.093 ms 6 - 217 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS7181 15.985 ms 6 - 6 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Gen 3 Mobile 24.839 ms 1 - 306 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Gen 1 Mobile 63.692 ms 3 - 304 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS8275 135.556 ms 2 - 210 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS8550 (Proxy) 36.294 ms 2 - 5 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8775P 417.782 ms 2 - 31 MB GPU
FFNet-40S TFLITE float Qualcomm® SA8650P 417.782 ms 2 - 31 MB GPU
FFNet-40S TFLITE float Qualcomm® SA8255P 417.782 ms 2 - 31 MB GPU
FFNet-40S TFLITE float Qualcomm® QCS8450 63.692 ms 3 - 304 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite Mobile 18.074 ms 2 - 239 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8295P 53.725 ms 3 - 217 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 12.638 ms 2 - 254 MB NPU
FFNet-40S TFLITE float Qualcomm® SA7255P 135.556 ms 2 - 210 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS9075 62.169 ms 0 - 56 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS8750 18.074 ms 2 - 239 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 5.449 ms 0 - 247 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 12.303 ms 1 - 249 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS6490 52.15 ms 1 - 23 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8275 20.059 ms 1 - 196 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8550 (Proxy) 7.452 ms 1 - 3 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8775P 175.554 ms 27 - 57 MB GPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8650P 175.554 ms 27 - 57 MB GPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8255P 175.554 ms 27 - 57 MB GPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8450 12.303 ms 1 - 249 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 2.924 ms 1 - 226 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 11.517 ms 1 - 215 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCM6690 102.752 ms 2 - 232 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS9075 9.579 ms 1 - 23 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA7255P 20.059 ms 1 - 196 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite Mobile 3.997 ms 0 - 212 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8295P 11.681 ms 1 - 200 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS7790 11.517 ms 1 - 215 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8750 3.997 ms 0 - 212 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