Hugging Face's logo

qualcomm
/
Segformer-Base

Image Segmentation
PyTorch
real_time
android
Model card Files
xet
 Community

Segformer-Base: Optimized for Qualcomm Devices

Segformer Base is a machine learning model that predicts masks and classes of objects in an image.

This is based on the implementation of Segformer-Base 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 w8a16 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 w8a16 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 Segformer-Base 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 Segformer-Base on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

  • Model checkpoint: nvidia/segformer-b0-finetuned-ade-512-512
  • Input resolution: 512x512
  • Number of output classes: 150
  • Number of parameters: 3.75M
  • Model size (float): 14.4 MB
  • Model size (w8a16): 4.57 MB
  • Model size (w8a8): 3.90 MB

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
Segformer-Base ONNX float Snapdragon® X2 Elite 13.523 ms 35 - 35 MB NPU
Segformer-Base ONNX float Snapdragon® X Elite 24.584 ms 35 - 35 MB NPU
Segformer-Base ONNX float Snapdragon® 8 Gen 3 Mobile 17.358 ms 25 - 241 MB NPU
Segformer-Base ONNX float Snapdragon® 8 Gen 1 Mobile 24.781 ms 13 - 232 MB NPU
Segformer-Base ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 23.739 ms 22 - 149 MB NPU
Segformer-Base ONNX float Qualcomm® QCS8450 24.781 ms 13 - 232 MB NPU
Segformer-Base ONNX float Snapdragon® 8 Elite Gen 5 Mobile 13.269 ms 21 - 211 MB NPU
Segformer-Base ONNX float Snapdragon® 8 Elite Mobile 14.413 ms 21 - 208 MB NPU
Segformer-Base ONNX float Qualcomm® Dragonwing™ IQ-9075 25.41 ms 23 - 26 MB NPU
Segformer-Base ONNX float Qualcomm® Dragonwing™ Q-8750 14.413 ms 21 - 208 MB NPU
Segformer-Base ONNX float Qualcomm® Dragonwing™ IQ-X7181 24.584 ms 35 - 35 MB NPU
Segformer-Base ONNX w8a16 Snapdragon® X2 Elite 5.62 ms 18 - 18 MB NPU
Segformer-Base ONNX w8a16 Snapdragon® X Elite 13.445 ms 17 - 17 MB NPU
Segformer-Base ONNX w8a16 Snapdragon® 8 Gen 3 Mobile 9.396 ms 14 - 247 MB NPU
Segformer-Base ONNX w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 13.313 ms 9 - 16 MB NPU
Segformer-Base ONNX w8a16 Snapdragon® 8 Elite Mobile 7.251 ms 14 - 217 MB NPU
Segformer-Base ONNX w8a16 Qualcomm® Dragonwing™ IQ-9075 14.461 ms 14 - 16 MB NPU
Segformer-Base ONNX w8a16 Snapdragon® 7 Gen 4 Mobile 17.071 ms 14 - 236 MB NPU
Segformer-Base ONNX w8a16 Snapdragon® 8 Elite Gen 5 Mobile 5.669 ms 13 - 221 MB NPU
Segformer-Base ONNX w8a16 Qualcomm® Dragonwing™ Q-6690 91.179 ms 13 - 244 MB NPU
Segformer-Base ONNX w8a16 Qualcomm® Dragonwing™ Q-7790 17.071 ms 14 - 236 MB NPU
Segformer-Base ONNX w8a16 Qualcomm® Dragonwing™ Q-8750 7.251 ms 14 - 217 MB NPU
Segformer-Base ONNX w8a16 Qualcomm® Dragonwing™ IQ-X7181 13.445 ms 17 - 17 MB NPU
Segformer-Base ONNX w8a8 Snapdragon® X2 Elite 1.949 ms 9 - 9 MB NPU
Segformer-Base ONNX w8a8 Snapdragon® X Elite 4.99 ms 5 - 5 MB NPU
Segformer-Base ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 3.227 ms 0 - 214 MB NPU
Segformer-Base ONNX w8a8 Snapdragon® 8 Gen 1 Mobile 7.033 ms 4 - 221 MB NPU
Segformer-Base ONNX w8a8 Qualcomm® Dragonwing™ QCS6490 19.805 ms 6 - 9 MB NPU
Segformer-Base ONNX w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 4.753 ms 5 - 12 MB NPU
Segformer-Base ONNX w8a8 Qualcomm® QCS8450 7.033 ms 4 - 221 MB NPU
Segformer-Base ONNX w8a8 Snapdragon® 8 Elite Mobile 2.387 ms 0 - 193 MB NPU
Segformer-Base ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 5.393 ms 9 - 206 MB NPU
Segformer-Base ONNX w8a8 Qualcomm® Dragonwing™ IQ-9075 5.116 ms 6 - 8 MB NPU
Segformer-Base ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 1.91 ms 0 - 196 MB NPU
Segformer-Base ONNX w8a8 Qualcomm® Dragonwing™ Q-6690 29.703 ms 7 - 236 MB NPU
Segformer-Base ONNX w8a8 Qualcomm® Dragonwing™ Q-7790 5.393 ms 9 - 206 MB NPU
Segformer-Base ONNX w8a8 Qualcomm® Dragonwing™ Q-8750 2.387 ms 0 - 193 MB NPU
Segformer-Base ONNX w8a8 Qualcomm® Dragonwing™ IQ-X7181 4.99 ms 5 - 5 MB NPU
Segformer-Base QNN_DLC float Snapdragon® X2 Elite 12.602 ms 3 - 3 MB NPU
Segformer-Base QNN_DLC float Snapdragon® X Elite 23.529 ms 3 - 3 MB NPU
Segformer-Base QNN_DLC float Snapdragon® 8 Gen 3 Mobile 16.421 ms 0 - 224 MB NPU
Segformer-Base QNN_DLC float Snapdragon® 8 Gen 1 Mobile 29.357 ms 3 - 226 MB NPU
Segformer-Base QNN_DLC float Qualcomm® QCS8275 48.882 ms 0 - 190 MB NPU
Segformer-Base QNN_DLC float Qualcomm® Dragonwing™ QCS8550 (Proxy) 22.717 ms 3 - 22 MB NPU
Segformer-Base QNN_DLC float Qualcomm® SA8775P 24.418 ms 0 - 190 MB NPU
Segformer-Base QNN_DLC float Qualcomm® SA8650P 24.418 ms 0 - 190 MB NPU
Segformer-Base QNN_DLC float Qualcomm® SA8255P 24.418 ms 0 - 190 MB NPU
Segformer-Base QNN_DLC float Qualcomm® QCS8450 29.357 ms 3 - 226 MB NPU
Segformer-Base QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 13.722 ms 3 - 192 MB NPU
Segformer-Base QNN_DLC float Snapdragon® 8 Elite Mobile 13.125 ms 0 - 197 MB NPU
Segformer-Base QNN_DLC float Qualcomm® Dragonwing™ IQ-9075 25.004 ms 3 - 17 MB NPU
Segformer-Base QNN_DLC float Qualcomm® SA7255P 48.882 ms 0 - 190 MB NPU
Segformer-Base QNN_DLC float Qualcomm® SA8295P 27.859 ms 0 - 191 MB NPU
Segformer-Base QNN_DLC float Qualcomm® Dragonwing™ Q-8750 13.125 ms 0 - 197 MB NPU
Segformer-Base QNN_DLC float Qualcomm® Dragonwing™ IQ-X7181 23.529 ms 3 - 3 MB NPU
Segformer-Base QNN_DLC w8a16 Snapdragon® X2 Elite 8.956 ms 2 - 2 MB NPU
Segformer-Base QNN_DLC w8a16 Snapdragon® X Elite 19.504 ms 2 - 2 MB NPU
Segformer-Base QNN_DLC w8a16 Snapdragon® 8 Gen 3 Mobile 15.2 ms 0 - 239 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® QCS8275 32.41 ms 2 - 204 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 20.429 ms 2 - 3 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® SA8775P 20.2 ms 2 - 206 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® SA8650P 20.2 ms 2 - 206 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® SA8255P 20.2 ms 2 - 206 MB NPU
Segformer-Base QNN_DLC w8a16 Snapdragon® 8 Elite Mobile 11.232 ms 2 - 208 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-9075 20.443 ms 2 - 9 MB NPU
Segformer-Base QNN_DLC w8a16 Snapdragon® 7 Gen 4 Mobile 34.593 ms 2 - 232 MB NPU
Segformer-Base QNN_DLC w8a16 Snapdragon® 8 Elite Gen 5 Mobile 9.75 ms 2 - 217 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-6690 115.575 ms 2 - 254 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-7790 34.593 ms 2 - 232 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® Dragonwing™ Q-8750 11.232 ms 2 - 208 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® Dragonwing™ IQ-X7181 19.504 ms 2 - 2 MB NPU
Segformer-Base QNN_DLC w8a16 Qualcomm® SA7255P 32.41 ms 2 - 204 MB NPU
Segformer-Base QNN_DLC w8a8 Snapdragon® X2 Elite 2.526 ms 1 - 1 MB NPU
Segformer-Base QNN_DLC w8a8 Snapdragon® X Elite 6.326 ms 1 - 1 MB NPU
Segformer-Base QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 3.941 ms 0 - 196 MB NPU
Segformer-Base QNN_DLC w8a8 Snapdragon® 8 Gen 1 Mobile 8.58 ms 0 - 199 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS6490 21.759 ms 3 - 7 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® QCS8275 10.872 ms 1 - 174 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 5.766 ms 0 - 18 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® SA8775P 6.419 ms 1 - 175 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® SA8650P 6.419 ms 1 - 175 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® SA8255P 6.419 ms 1 - 175 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® QCS8450 8.58 ms 0 - 199 MB NPU
Segformer-Base QNN_DLC w8a8 Snapdragon® 8 Elite Mobile 2.854 ms 0 - 173 MB NPU
Segformer-Base QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 6.832 ms 0 - 178 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-9075 6.131 ms 1 - 5 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® SA7255P 10.872 ms 1 - 174 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® SA8295P 7.788 ms 0 - 172 MB NPU
Segformer-Base QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 2.239 ms 1 - 183 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-6690 28.937 ms 1 - 188 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-7790 6.832 ms 0 - 178 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® Dragonwing™ Q-8750 2.854 ms 0 - 173 MB NPU
Segformer-Base QNN_DLC w8a8 Qualcomm® Dragonwing™ IQ-X7181 6.326 ms 1 - 1 MB NPU
Segformer-Base TFLITE float Snapdragon® 8 Gen 3 Mobile 16.51 ms 8 - 236 MB NPU
Segformer-Base TFLITE float Snapdragon® 8 Gen 1 Mobile 29.265 ms 9 - 233 MB NPU
Segformer-Base TFLITE float Qualcomm® QCS8275 48.989 ms 9 - 200 MB NPU
Segformer-Base TFLITE float Qualcomm® Dragonwing™ QCS8550 (Proxy) 22.681 ms 9 - 104 MB NPU
Segformer-Base TFLITE float Qualcomm® SA8775P 24.499 ms 0 - 192 MB NPU
Segformer-Base TFLITE float Qualcomm® SA8650P 24.499 ms 0 - 192 MB NPU
Segformer-Base TFLITE float Qualcomm® SA8255P 24.499 ms 0 - 192 MB NPU
Segformer-Base TFLITE float Qualcomm® QCS8450 29.265 ms 9 - 233 MB NPU
Segformer-Base TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 13.679 ms 10 - 201 MB NPU
Segformer-Base TFLITE float Snapdragon® 8 Elite Mobile 13.18 ms 9 - 203 MB NPU
Segformer-Base TFLITE float Qualcomm® Dragonwing™ IQ-9075 25.671 ms 8 - 30 MB NPU
Segformer-Base TFLITE float Qualcomm® SA7255P 48.989 ms 9 - 200 MB NPU
Segformer-Base TFLITE float Qualcomm® SA8295P 27.81 ms 9 - 202 MB NPU
Segformer-Base TFLITE float Qualcomm® Dragonwing™ Q-8750 13.18 ms 9 - 203 MB NPU
Segformer-Base TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 7.104 ms 2 - 213 MB NPU
Segformer-Base TFLITE w8a8 Snapdragon® 8 Gen 1 Mobile 12.1 ms 2 - 210 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® Dragonwing™ QCS6490 134.565 ms 15 - 50 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® QCS8275 18.268 ms 2 - 178 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® Dragonwing™ QCS8550 (Proxy) 10.127 ms 2 - 5 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® SA8775P 10.881 ms 2 - 183 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® SA8650P 10.881 ms 2 - 183 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® SA8255P 10.881 ms 2 - 183 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® QCS8450 12.1 ms 2 - 210 MB NPU
Segformer-Base TFLITE w8a8 Snapdragon® 8 Elite Mobile 5.712 ms 3 - 179 MB NPU
Segformer-Base TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 38.939 ms 15 - 73 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® Dragonwing™ IQ-9075 10.617 ms 2 - 12 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® SA7255P 18.268 ms 2 - 178 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® SA8295P 12.85 ms 2 - 187 MB NPU
Segformer-Base TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 4.553 ms 1 - 179 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® Dragonwing™ Q-6690 96.343 ms 13 - 187 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® Dragonwing™ Q-7790 38.939 ms 15 - 73 MB NPU
Segformer-Base TFLITE w8a8 Qualcomm® Dragonwing™ Q-8750 5.712 ms 3 - 179 MB NPU

License

  • The license for the original implementation of Segformer-Base 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/Segformer-Base