Hugging Face's logo

qualcomm
/
DeepLabXception

Image Segmentation
PyTorch
bu_auto
android
Model card Files
xet
 Community
DeepLabXception / README.md
qaihm-bot's picture
qaihm-bot
v0.47.0
98f620f verified
preview code
|
raw
history blame
10.7 kB
metadata 
library_name: pytorch
license: other
tags:
  - bu_auto
  - android
pipeline_tag: image-segmentation

DeepLabXception: Optimized for Qualcomm Devices

DeepLabXception is a semantic segmentation model supporting multiple backbones like ResNet-101 and Xception, with flexible dataset compatibility including COCO, VOC, and Cityscapes.

This is based on the implementation of DeepLabXception 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 1.24.1 Download
ONNX w8a8 Universal QAIRT 2.42, ONNX Runtime 1.24.1 Download
QNN_DLC float Universal QAIRT 2.43 Download
QNN_DLC w8a8 Universal QAIRT 2.43 Download
TFLITE float Universal QAIRT 2.43, TFLite 2.17.0 Download
TFLITE w8a8 Universal QAIRT 2.43, TFLite 2.17.0 Download

For more device-specific assets and performance metrics, visit DeepLabXception 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 DeepLabXception on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

  • Model checkpoint: COCO_WITH_VOC_LABELS_V1
  • Input resolution: 480x520
  • Number of output classes: 21
  • Number of parameters: 41.26M
  • Model size (float): 158 MB

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
DeepLabXception ONNX float Snapdragon® X Elite 22.73 ms 85 - 85 MB NPU
DeepLabXception ONNX float Snapdragon® 8 Gen 3 Mobile 16.834 ms 0 - 366 MB NPU
DeepLabXception ONNX float Qualcomm® QCS8550 (Proxy) 22.198 ms 0 - 95 MB NPU
DeepLabXception ONNX float Qualcomm® QCS9075 31.778 ms 3 - 8 MB NPU
DeepLabXception ONNX float Snapdragon® 8 Elite For Galaxy Mobile 13.178 ms 1 - 252 MB NPU
DeepLabXception ONNX float Snapdragon® 8 Elite Gen 5 Mobile 8.733 ms 3 - 255 MB NPU
DeepLabXception ONNX float Snapdragon® X2 Elite 10.394 ms 86 - 86 MB NPU
DeepLabXception ONNX w8a8 Snapdragon® X Elite 8.003 ms 44 - 44 MB NPU
DeepLabXception ONNX w8a8 Snapdragon® 8 Gen 3 Mobile 5.259 ms 0 - 388 MB NPU
DeepLabXception ONNX w8a8 Qualcomm® QCS6490 316.797 ms 92 - 118 MB CPU
DeepLabXception ONNX w8a8 Qualcomm® QCS8550 (Proxy) 7.474 ms 0 - 87 MB NPU
DeepLabXception ONNX w8a8 Qualcomm® QCS9075 7.614 ms 0 - 4 MB NPU
DeepLabXception ONNX w8a8 Qualcomm® QCM6690 244.493 ms 90 - 102 MB CPU
DeepLabXception ONNX w8a8 Snapdragon® 8 Elite For Galaxy Mobile 4.216 ms 1 - 289 MB NPU
DeepLabXception ONNX w8a8 Snapdragon® 7 Gen 4 Mobile 229.605 ms 89 - 102 MB CPU
DeepLabXception ONNX w8a8 Snapdragon® 8 Elite Gen 5 Mobile 3.06 ms 0 - 292 MB NPU
DeepLabXception ONNX w8a8 Snapdragon® X2 Elite 3.353 ms 44 - 44 MB NPU
DeepLabXception QNN_DLC float Snapdragon® X Elite 22.747 ms 3 - 3 MB NPU
DeepLabXception QNN_DLC float Snapdragon® 8 Gen 3 Mobile 16.101 ms 3 - 369 MB NPU
DeepLabXception QNN_DLC float Qualcomm® QCS8275 (Proxy) 113.153 ms 1 - 244 MB NPU
DeepLabXception QNN_DLC float Qualcomm® QCS8550 (Proxy) 22.063 ms 3 - 5 MB NPU
DeepLabXception QNN_DLC float Qualcomm® SA8775P 30.583 ms 1 - 244 MB NPU
DeepLabXception QNN_DLC float Qualcomm® QCS9075 35.913 ms 5 - 10 MB NPU
DeepLabXception QNN_DLC float Qualcomm® QCS8450 (Proxy) 46.424 ms 0 - 362 MB NPU
DeepLabXception QNN_DLC float Qualcomm® SA7255P 113.153 ms 1 - 244 MB NPU
DeepLabXception QNN_DLC float Qualcomm® SA8295P 38.871 ms 0 - 242 MB NPU
DeepLabXception QNN_DLC float Snapdragon® 8 Elite For Galaxy Mobile 12.144 ms 3 - 252 MB NPU
DeepLabXception QNN_DLC float Snapdragon® 8 Elite Gen 5 Mobile 9.212 ms 3 - 256 MB NPU
DeepLabXception QNN_DLC float Snapdragon® X2 Elite 11.197 ms 3 - 3 MB NPU
DeepLabXception QNN_DLC w8a8 Snapdragon® X Elite 8.555 ms 1 - 1 MB NPU
DeepLabXception QNN_DLC w8a8 Snapdragon® 8 Gen 3 Mobile 5.549 ms 1 - 375 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® QCS6490 33.63 ms 1 - 3 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® QCS8275 (Proxy) 20.332 ms 1 - 281 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® QCS8550 (Proxy) 7.917 ms 1 - 3 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® SA8775P 8.102 ms 1 - 284 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® QCS9075 8.586 ms 3 - 5 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® QCM6690 118.645 ms 1 - 303 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® QCS8450 (Proxy) 13.841 ms 1 - 368 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® SA7255P 20.332 ms 1 - 281 MB NPU
DeepLabXception QNN_DLC w8a8 Qualcomm® SA8295P 12.677 ms 1 - 277 MB NPU
DeepLabXception QNN_DLC w8a8 Snapdragon® 8 Elite For Galaxy Mobile 4.359 ms 0 - 275 MB NPU
DeepLabXception QNN_DLC w8a8 Snapdragon® 7 Gen 4 Mobile 12.88 ms 1 - 277 MB NPU
DeepLabXception QNN_DLC w8a8 Snapdragon® 8 Elite Gen 5 Mobile 3.259 ms 1 - 274 MB NPU
DeepLabXception QNN_DLC w8a8 Snapdragon® X2 Elite 3.887 ms 1 - 1 MB NPU
DeepLabXception TFLITE float Snapdragon® 8 Gen 3 Mobile 16.164 ms 0 - 482 MB NPU
DeepLabXception TFLITE float Qualcomm® QCS8275 (Proxy) 113.255 ms 0 - 358 MB NPU
DeepLabXception TFLITE float Qualcomm® QCS8550 (Proxy) 21.326 ms 0 - 3 MB NPU
DeepLabXception TFLITE float Qualcomm® SA8775P 30.569 ms 0 - 359 MB NPU
DeepLabXception TFLITE float Qualcomm® QCS9075 35.093 ms 0 - 93 MB NPU
DeepLabXception TFLITE float Qualcomm® QCS8450 (Proxy) 45.599 ms 0 - 470 MB NPU
DeepLabXception TFLITE float Qualcomm® SA7255P 113.255 ms 0 - 358 MB NPU
DeepLabXception TFLITE float Qualcomm® SA8295P 38.658 ms 0 - 349 MB NPU
DeepLabXception TFLITE float Snapdragon® 8 Elite For Galaxy Mobile 12.201 ms 0 - 367 MB NPU
DeepLabXception TFLITE float Snapdragon® 8 Elite Gen 5 Mobile 9.202 ms 0 - 364 MB NPU
DeepLabXception TFLITE w8a8 Snapdragon® 8 Gen 3 Mobile 4.961 ms 0 - 372 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® QCS6490 32.06 ms 0 - 58 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® QCS8275 (Proxy) 18.898 ms 0 - 285 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® QCS8550 (Proxy) 6.94 ms 0 - 3 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® SA8775P 7.239 ms 0 - 285 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® QCS9075 7.444 ms 0 - 47 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® QCM6690 114.998 ms 0 - 295 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® QCS8450 (Proxy) 11.873 ms 0 - 360 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® SA7255P 18.898 ms 0 - 285 MB NPU
DeepLabXception TFLITE w8a8 Qualcomm® SA8295P 11.545 ms 0 - 274 MB NPU
DeepLabXception TFLITE w8a8 Snapdragon® 8 Elite For Galaxy Mobile 3.936 ms 0 - 274 MB NPU
DeepLabXception TFLITE w8a8 Snapdragon® 7 Gen 4 Mobile 11.295 ms 0 - 270 MB NPU
DeepLabXception TFLITE w8a8 Snapdragon® 8 Elite Gen 5 Mobile 2.896 ms 0 - 267 MB NPU

License

  • The license for the original implementation of DeepLabXception can be found here.

References

Community