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 2.42, ONNX Runtime 1.24.3 Download
ONNX w8a8 Universal QAIRT 2.42, ONNX Runtime 1.24.3 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® 8 Elite Gen 5 Mobile 12.36 ms 30 - 253 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Elite Mobile 16.005 ms 6 - 208 MB NPU
FFNet-40S ONNX float Snapdragon® X2 Elite 13.434 ms 22 - 22 MB NPU
FFNet-40S ONNX float Snapdragon® X Elite 31.477 ms 24 - 24 MB NPU
FFNet-40S ONNX float Snapdragon® X Elite 31.477 ms 24 - 24 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Gen 3 Mobile 22.44 ms 30 - 303 MB NPU
FFNet-40S ONNX float Qualcomm® QCS8550 (Proxy) 32.108 ms 24 - 55 MB NPU
FFNet-40S ONNX float Snapdragon® 8 Elite For Galaxy Mobile 16.005 ms 6 - 208 MB NPU
FFNet-40S ONNX float Qualcomm® QCS9075 48.076 ms 24 - 27 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 6.553 ms 2 - 197 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite Mobile 7.732 ms 1 - 193 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X2 Elite 6.925 ms 9 - 9 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X Elite 10.477 ms 8 - 8 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® X Elite 10.477 ms 8 - 8 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 6.573 ms 7 - 250 MB NPU
FFNet-40S ONNX w8a8 Qualcomm® QCS6490 358.87 ms 204 - 239 MB CPU
FFNet-40S ONNX w8a8 Qualcomm® QCS8550 (Proxy) 9.827 ms 6 - 10 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 359.709 ms 166 - 175 MB CPU
FFNet-40S ONNX w8a8 Qualcomm® QCM6690 365.5 ms 163 - 171 MB CPU
FFNet-40S ONNX w8a8 Qualcomm® QCS9075 13.231 ms 6 - 9 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 8 Elite For Galaxy Mobile 7.732 ms 1 - 193 MB NPU
FFNet-40S ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 359.709 ms 166 - 175 MB CPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 12.634 ms 24 - 263 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite Mobile 18.032 ms 20 - 245 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X2 Elite 14.574 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X Elite 37.479 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® X Elite 37.479 ms 24 - 24 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Gen 3 Mobile 24.798 ms 22 - 290 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8550 (Proxy) 36.14 ms 28 - 40 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8775P 48.917 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8775P 48.917 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8775P 48.917 ms 24 - 219 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA7255P 135.538 ms 24 - 218 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS8450 (Proxy) 63.693 ms 0 - 265 MB NPU
FFNet-40S QNN_DLC float Qualcomm® SA8295P 53.715 ms 24 - 224 MB NPU
FFNet-40S QNN_DLC float Snapdragon® 8 Elite For Galaxy Mobile 18.032 ms 20 - 245 MB NPU
FFNet-40S QNN_DLC float Qualcomm® QCS9075 62.078 ms 24 - 52 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 5.229 ms 6 - 230 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 7.1 ms 6 - 219 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X2 Elite 6.069 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X Elite 15.922 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® X Elite 15.922 ms 6 - 6 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 10.499 ms 6 - 250 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS6490 66.677 ms 3 - 11 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8550 (Proxy) 15.08 ms 6 - 8 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8775P 15.674 ms 6 - 203 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8775P 15.674 ms 6 - 203 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8775P 15.674 ms 6 - 203 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 18.988 ms 6 - 223 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCM6690 134.22 ms 6 - 238 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS9075 18.539 ms 6 - 14 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 8 Elite For Galaxy Mobile 7.1 ms 6 - 219 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® QCS8450 (Proxy) 21.627 ms 6 - 253 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA7255P 32.902 ms 6 - 202 MB NPU
FFNet-40S QNN_DLC w8a8 Qualcomm® SA8295P 20.046 ms 6 - 206 MB NPU
FFNet-40S QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 18.988 ms 6 - 223 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 12.611 ms 2 - 252 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite Mobile 18.115 ms 2 - 241 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Gen 3 Mobile 25.046 ms 1 - 305 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS8550 (Proxy) 36.108 ms 0 - 282 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8775P 48.966 ms 2 - 211 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8775P 48.966 ms 2 - 211 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8775P 48.966 ms 2 - 211 MB NPU
FFNet-40S TFLITE float Qualcomm® SA7255P 135.606 ms 3 - 210 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS8450 (Proxy) 64.146 ms 2 - 304 MB NPU
FFNet-40S TFLITE float Qualcomm® SA8295P 53.712 ms 2 - 216 MB NPU
FFNet-40S TFLITE float Snapdragon® 8 Elite For Galaxy Mobile 18.115 ms 2 - 241 MB NPU
FFNet-40S TFLITE float Qualcomm® QCS9075 62.263 ms 0 - 56 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 2.913 ms 1 - 224 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite Mobile 3.987 ms 0 - 210 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 5.463 ms 0 - 248 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS6490 52.037 ms 1 - 23 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8550 (Proxy) 7.469 ms 1 - 12 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8775P 8.257 ms 0 - 196 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8775P 8.257 ms 0 - 196 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8775P 8.257 ms 0 - 196 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 11.461 ms 1 - 216 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCM6690 103.122 ms 0 - 230 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS9075 9.579 ms 1 - 23 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 8 Elite For Galaxy Mobile 3.987 ms 0 - 210 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® QCS8450 (Proxy) 12.316 ms 1 - 250 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA7255P 20.049 ms 1 - 196 MB NPU
FFNet-40S TFLITE w8a8 Qualcomm® SA8295P 11.678 ms 1 - 198 MB NPU
FFNet-40S TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 11.461 ms 1 - 216 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