---
library_name: pytorch
license: other
tags:
- bu_auto
- android
pipeline_tag: image-segmentation

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deeplab_xception/web-assets/model_demo.png)

# 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](https://github.com/LikeLy-Journey/SegmenTron).
This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/deeplab_xception) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).

Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) 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](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deeplab_xception/releases/v0.53.1/deeplab_xception-onnx-float.zip)
| ONNX | w8a8 | Universal | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deeplab_xception/releases/v0.53.1/deeplab_xception-onnx-w8a8.zip)
| QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deeplab_xception/releases/v0.53.1/deeplab_xception-qnn_dlc-float.zip)
| QNN_DLC | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deeplab_xception/releases/v0.53.1/deeplab_xception-qnn_dlc-w8a8.zip)
| TFLITE | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deeplab_xception/releases/v0.53.1/deeplab_xception-tflite-float.zip)
| TFLITE | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deeplab_xception/releases/v0.53.1/deeplab_xception-tflite-w8a8.zip)

For more device-specific assets and performance metrics, visit **[DeepLabXception on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/deeplab_xception)**.


### Option 2: Export with Custom Configurations

Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/deeplab_xception) 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](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/deeplab_xception) 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® 8 Elite Gen 5 Mobile | 8.903 ms | 3 - 255 MB | NPU
| DeepLabXception | ONNX | float | Snapdragon® 8 Elite Mobile | 13.513 ms | 1 - 253 MB | NPU
| DeepLabXception | ONNX | float | Snapdragon® X2 Elite | 10.393 ms | 86 - 86 MB | NPU
| DeepLabXception | ONNX | float | Snapdragon® X Elite | 22.933 ms | 85 - 85 MB | NPU
| DeepLabXception | ONNX | float | Snapdragon® X Elite | 22.933 ms | 85 - 85 MB | NPU
| DeepLabXception | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 17.152 ms | 0 - 366 MB | NPU
| DeepLabXception | ONNX | float | Qualcomm® QCS8550 (Proxy) | 22.673 ms | 3 - 6 MB | NPU
| DeepLabXception | ONNX | float | Qualcomm® QCS9075 | 33.606 ms | 3 - 8 MB | NPU
| DeepLabXception | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 13.513 ms | 1 - 253 MB | NPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 3.056 ms | 0 - 293 MB | NPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® 8 Elite Mobile | 4.23 ms | 0 - 288 MB | NPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® X2 Elite | 3.356 ms | 44 - 44 MB | NPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® X Elite | 8.017 ms | 44 - 44 MB | NPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® X Elite | 8.017 ms | 44 - 44 MB | NPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 5.274 ms | 0 - 389 MB | NPU
| DeepLabXception | ONNX | w8a8 | Qualcomm® QCS6490 | 309.047 ms | 93 - 117 MB | CPU
| DeepLabXception | ONNX | w8a8 | Qualcomm® QCS8550 (Proxy) | 7.442 ms | 0 - 100 MB | NPU
| DeepLabXception | ONNX | w8a8 | Qualcomm® QCS9075 | 7.6 ms | 0 - 4 MB | NPU
| DeepLabXception | ONNX | w8a8 | Qualcomm® QCM6690 | 243.888 ms | 90 - 103 MB | CPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 4.23 ms | 0 - 288 MB | NPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 229.65 ms | 87 - 101 MB | CPU
| DeepLabXception | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 229.65 ms | 87 - 101 MB | CPU
| DeepLabXception | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 9.544 ms | 3 - 257 MB | NPU
| DeepLabXception | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 12.278 ms | 0 - 251 MB | NPU
| DeepLabXception | QNN_DLC | float | Snapdragon® X2 Elite | 11.525 ms | 3 - 3 MB | NPU
| DeepLabXception | QNN_DLC | float | Snapdragon® X Elite | 22.879 ms | 3 - 3 MB | NPU
| DeepLabXception | QNN_DLC | float | Snapdragon® X Elite | 22.879 ms | 3 - 3 MB | NPU
| DeepLabXception | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 16.137 ms | 0 - 353 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 112.951 ms | 1 - 244 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 22.031 ms | 3 - 5 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® SA8775P | 30.627 ms | 1 - 244 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® SA8775P | 30.627 ms | 1 - 244 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® SA8775P | 30.627 ms | 1 - 244 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® QCS9075 | 35.833 ms | 3 - 8 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 45.907 ms | 0 - 351 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® SA7255P | 112.951 ms | 1 - 244 MB | NPU
| DeepLabXception | QNN_DLC | float | Qualcomm® SA8295P | 39.155 ms | 2 - 241 MB | NPU
| DeepLabXception | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 12.278 ms | 0 - 251 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 3.25 ms | 1 - 280 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® 8 Elite Mobile | 4.375 ms | 1 - 289 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 3.88 ms | 1 - 1 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® X Elite | 8.519 ms | 1 - 1 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® X Elite | 8.519 ms | 1 - 1 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 5.581 ms | 1 - 373 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® QCS6490 | 33.678 ms | 3 - 5 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® QCS8275 (Proxy) | 20.363 ms | 1 - 282 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® QCS8550 (Proxy) | 7.873 ms | 1 - 3 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® SA8775P | 8.066 ms | 1 - 285 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® SA8775P | 8.066 ms | 1 - 285 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® SA8775P | 8.066 ms | 1 - 285 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® QCS9075 | 8.636 ms | 1 - 3 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® QCM6690 | 93.567 ms | 1 - 304 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® QCS8450 (Proxy) | 14.3 ms | 1 - 366 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® SA7255P | 20.363 ms | 1 - 282 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Qualcomm® SA8295P | 12.707 ms | 1 - 279 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 4.375 ms | 1 - 289 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 12.168 ms | 1 - 277 MB | NPU
| DeepLabXception | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 12.168 ms | 1 - 277 MB | NPU
| DeepLabXception | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 9.193 ms | 0 - 356 MB | NPU
| DeepLabXception | TFLITE | float | Snapdragon® 8 Elite Mobile | 12.379 ms | 0 - 355 MB | NPU
| DeepLabXception | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 16.177 ms | 0 - 455 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 113.155 ms | 0 - 344 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 21.918 ms | 0 - 3 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® SA8775P | 30.651 ms | 0 - 343 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® SA8775P | 30.651 ms | 0 - 343 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® SA8775P | 30.651 ms | 0 - 343 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® QCS9075 | 35.101 ms | 0 - 93 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 46.117 ms | 0 - 452 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® SA7255P | 113.155 ms | 0 - 344 MB | NPU
| DeepLabXception | TFLITE | float | Qualcomm® SA8295P | 38.739 ms | 0 - 336 MB | NPU
| DeepLabXception | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 12.379 ms | 0 - 355 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 2.88 ms | 0 - 283 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Snapdragon® 8 Elite Mobile | 3.939 ms | 0 - 270 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 4.96 ms | 0 - 367 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® QCS6490 | 31.95 ms | 0 - 58 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® QCS8275 (Proxy) | 19.018 ms | 0 - 283 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® QCS8550 (Proxy) | 6.976 ms | 0 - 4 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® SA8775P | 7.281 ms | 0 - 285 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® SA8775P | 7.281 ms | 0 - 285 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® SA8775P | 7.281 ms | 0 - 285 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® QCS9075 | 7.486 ms | 0 - 47 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® QCM6690 | 123.747 ms | 0 - 292 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® QCS8450 (Proxy) | 12.283 ms | 0 - 353 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® SA7255P | 19.018 ms | 0 - 283 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Qualcomm® SA8295P | 11.647 ms | 0 - 275 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 3.939 ms | 0 - 270 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 11.223 ms | 0 - 265 MB | NPU
| DeepLabXception | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 11.223 ms | 0 - 265 MB | NPU

## License
* The license for the original implementation of DeepLabXception can be found
  [here](https://github.com/pytorch/vision/blob/main/LICENSE).

## References
* [Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation](https://arxiv.org/abs/1802.02611)
* [Source Model Implementation](https://github.com/LikeLy-Journey/SegmenTron)

## Community
* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
* For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).