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 2.37, ONNX Runtime 1.23.0	Download
ONNX	w8a8	Universal	QAIRT 2.37, ONNX Runtime 1.23.0	Download
QNN_DLC	float	Universal	QAIRT 2.42	Download
QNN_DLC	w8a8	Universal	QAIRT 2.42	Download
TFLITE	float	Universal	QAIRT 2.42, TFLite 2.17.0	Download
TFLITE	w8a8	Universal	QAIRT 2.42, 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	21.791 ms	86 - 86 MB	NPU
DeepLabXception	ONNX	float	Snapdragon® 8 Gen 3 Mobile	15.806 ms	3 - 304 MB	NPU
DeepLabXception	ONNX	float	Qualcomm® QCS8550 (Proxy)	20.959 ms	0 - 96 MB	NPU
DeepLabXception	ONNX	float	Qualcomm® QCS9075	32.676 ms	3 - 8 MB	NPU
DeepLabXception	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	12.183 ms	2 - 185 MB	NPU
DeepLabXception	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	9.236 ms	0 - 198 MB	NPU
DeepLabXception	ONNX	w8a8	Snapdragon® X Elite	8.003 ms	44 - 44 MB	NPU
DeepLabXception	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	5.35 ms	0 - 337 MB	NPU
DeepLabXception	ONNX	w8a8	Qualcomm® QCS6490	310.74 ms	89 - 122 MB	CPU
DeepLabXception	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	7.601 ms	0 - 51 MB	NPU
DeepLabXception	ONNX	w8a8	Qualcomm® QCS9075	7.881 ms	0 - 4 MB	NPU
DeepLabXception	ONNX	w8a8	Qualcomm® QCM6690	242.805 ms	57 - 71 MB	CPU
DeepLabXception	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	4.337 ms	0 - 233 MB	NPU
DeepLabXception	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	227.823 ms	91 - 103 MB	CPU
DeepLabXception	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	3.299 ms	0 - 237 MB	NPU
DeepLabXception	QNN_DLC	float	Snapdragon® X Elite	22.787 ms	3 - 3 MB	NPU
DeepLabXception	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	16.092 ms	0 - 366 MB	NPU
DeepLabXception	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	113.009 ms	1 - 245 MB	NPU
DeepLabXception	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	22.013 ms	3 - 6 MB	NPU
DeepLabXception	QNN_DLC	float	Qualcomm® SA8775P	30.552 ms	2 - 246 MB	NPU
DeepLabXception	QNN_DLC	float	Qualcomm® QCS9075	35.61 ms	5 - 10 MB	NPU
DeepLabXception	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	46.253 ms	0 - 361 MB	NPU
DeepLabXception	QNN_DLC	float	Qualcomm® SA7255P	113.009 ms	1 - 245 MB	NPU
DeepLabXception	QNN_DLC	float	Qualcomm® SA8295P	38.823 ms	0 - 242 MB	NPU
DeepLabXception	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	12.132 ms	0 - 249 MB	NPU
DeepLabXception	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	9.141 ms	3 - 258 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Snapdragon® X Elite	8.561 ms	1 - 1 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	5.614 ms	1 - 373 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® QCS6490	33.905 ms	1 - 3 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	20.273 ms	1 - 283 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	7.903 ms	1 - 3 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® SA8775P	8.061 ms	1 - 285 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® QCS9075	8.573 ms	1 - 3 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® QCM6690	121.023 ms	1 - 307 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	14.214 ms	1 - 367 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® SA7255P	20.273 ms	1 - 283 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Qualcomm® SA8295P	12.638 ms	1 - 275 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	4.411 ms	1 - 273 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	13.07 ms	1 - 277 MB	NPU
DeepLabXception	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	3.297 ms	1 - 274 MB	NPU
DeepLabXception	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	16.101 ms	0 - 483 MB	NPU
DeepLabXception	TFLITE	float	Qualcomm® QCS8275 (Proxy)	113.095 ms	0 - 358 MB	NPU
DeepLabXception	TFLITE	float	Qualcomm® QCS8550 (Proxy)	21.319 ms	0 - 3 MB	NPU
DeepLabXception	TFLITE	float	Qualcomm® SA8775P	30.518 ms	0 - 358 MB	NPU
DeepLabXception	TFLITE	float	Qualcomm® QCS9075	35.068 ms	0 - 93 MB	NPU
DeepLabXception	TFLITE	float	Qualcomm® QCS8450 (Proxy)	46.083 ms	0 - 471 MB	NPU
DeepLabXception	TFLITE	float	Qualcomm® SA7255P	113.095 ms	0 - 358 MB	NPU
DeepLabXception	TFLITE	float	Qualcomm® SA8295P	38.54 ms	0 - 349 MB	NPU
DeepLabXception	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	12.23 ms	0 - 369 MB	NPU
DeepLabXception	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	9.127 ms	0 - 370 MB	NPU
DeepLabXception	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	5.024 ms	0 - 373 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® QCS6490	31.912 ms	0 - 58 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	19.046 ms	0 - 282 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	6.974 ms	0 - 3 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® SA8775P	7.236 ms	0 - 284 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® QCS9075	7.448 ms	0 - 47 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® QCM6690	122.599 ms	0 - 294 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	12.47 ms	0 - 362 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® SA7255P	19.046 ms	0 - 282 MB	NPU
DeepLabXception	TFLITE	w8a8	Qualcomm® SA8295P	11.577 ms	0 - 275 MB	NPU
DeepLabXception	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	3.939 ms	0 - 274 MB	NPU
DeepLabXception	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	11.535 ms	0 - 266 MB	NPU
DeepLabXception	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	2.911 ms	0 - 269 MB	NPU

License

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

References

Community

Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
For questions or feedback please reach out to us.

Downloads last month: -; Downloads are not tracked for this model. How to track

Paper for qualcomm/DeepLabXception

Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation

Paper • 1802.02611 • Published Feb 7, 2018