DDRNet23-Slim: Optimized for Qualcomm Devices

DDRNet23Slim is a machine learning model that segments an image into semantic classes, specifically designed for road-based scenes. It is designed for the application of self-driving cars.

This is based on the implementation of DDRNet23-Slim 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 1.27.1	Download
ONNX	w8a8	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a8	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit DDRNet23-Slim 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 DDRNet23-Slim on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

Model checkpoint: DDRNet23s_imagenet.pth
Inference latency: RealTime
Input resolution: 2048x1024
Number of output classes: 19
Number of parameters: 6.13M
Model size (float): 21.7 MB
Model size (w8a8): 6.11 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
DDRNet23-Slim	ONNX	float	Snapdragon® X2 Elite	11.675 ms	24 - 24 MB	NPU
DDRNet23-Slim	ONNX	float	Snapdragon® X Elite	27.809 ms	24 - 24 MB	NPU
DDRNet23-Slim	ONNX	float	Snapdragon® 8 Gen 3 Mobile	19.922 ms	28 - 297 MB	NPU
DDRNet23-Slim	ONNX	float	Snapdragon® 8 Gen 1 Mobile	46.111 ms	24 - 296 MB	NPU
DDRNet23-Slim	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	28.564 ms	22 - 24 MB	NPU
DDRNet23-Slim	ONNX	float	Qualcomm® QCS8450	46.111 ms	24 - 296 MB	NPU
DDRNet23-Slim	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	35.702 ms	24 - 51 MB	NPU
DDRNet23-Slim	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	27.809 ms	24 - 24 MB	NPU
DDRNet23-Slim	ONNX	float	Qualcomm® Dragonwing™ Q-8750	15.278 ms	8 - 194 MB	NPU
DDRNet23-Slim	ONNX	float	Snapdragon® 8 Elite Mobile	15.278 ms	8 - 194 MB	NPU
DDRNet23-Slim	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	11.404 ms	5 - 219 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Snapdragon® X2 Elite	43.793 ms	7 - 7 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Snapdragon® X Elite	56.063 ms	7 - 7 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	39.526 ms	7 - 251 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	55.093 ms	8 - 253 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Qualcomm® Dragonwing™ QCS6490	153.105 ms	6 - 9 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	53.361 ms	6 - 16 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Qualcomm® QCS8450	55.093 ms	8 - 253 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-9075	52.716 ms	6 - 9 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Qualcomm® Dragonwing™ IQ-X7181	56.063 ms	7 - 7 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Qualcomm® Dragonwing™ Q-8750	41.084 ms	1 - 201 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Snapdragon® 8 Elite Mobile	41.084 ms	1 - 201 MB	NPU
DDRNet23-Slim	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	43.406 ms	1 - 203 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Snapdragon® X2 Elite	11.881 ms	24 - 24 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Snapdragon® X Elite	34.27 ms	24 - 24 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	22.31 ms	17 - 290 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	66.591 ms	24 - 300 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8275	98.418 ms	24 - 220 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	33.034 ms	24 - 27 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® SA8775P	40.438 ms	24 - 221 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® SA8650P	40.438 ms	24 - 221 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® SA8255P	40.438 ms	24 - 221 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® QCS8450	66.591 ms	24 - 300 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	44.535 ms	24 - 52 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	34.27 ms	24 - 24 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	15.427 ms	15 - 235 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® SA7255P	98.418 ms	24 - 220 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Qualcomm® SA8295P	43.525 ms	24 - 229 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Snapdragon® 8 Elite Mobile	15.427 ms	15 - 235 MB	NPU
DDRNet23-Slim	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	10.512 ms	0 - 238 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Snapdragon® X2 Elite	29.371 ms	6 - 6 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Snapdragon® X Elite	58.901 ms	6 - 6 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	41.945 ms	6 - 252 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	80.283 ms	7 - 253 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS8275	107.783 ms	6 - 205 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	56.454 ms	6 - 8 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® SA8775P	56.87 ms	6 - 206 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® SA8650P	56.87 ms	6 - 206 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® SA8255P	56.87 ms	6 - 206 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® QCS8450	80.283 ms	7 - 253 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-9075	57.37 ms	6 - 14 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® Dragonwing™ IQ-X7181	58.901 ms	6 - 6 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® Dragonwing™ Q-8750	39.901 ms	6 - 222 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® SA7255P	107.783 ms	6 - 205 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Qualcomm® SA8295P	66.362 ms	6 - 208 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	39.901 ms	6 - 222 MB	NPU
DDRNet23-Slim	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	28.729 ms	6 - 240 MB	NPU
DDRNet23-Slim	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	22.341 ms	2 - 284 MB	NPU
DDRNet23-Slim	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	66.422 ms	3 - 286 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® Dragonwing™ QCS8275	98.317 ms	2 - 204 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	33.126 ms	2 - 5 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® SA8775P	40.458 ms	2 - 205 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® SA8650P	40.458 ms	2 - 205 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® SA8255P	40.458 ms	2 - 205 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® QCS8450	66.422 ms	3 - 286 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	41.244 ms	0 - 41 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	15.633 ms	1 - 229 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® SA7255P	98.317 ms	2 - 204 MB	NPU
DDRNet23-Slim	TFLITE	float	Qualcomm® SA8295P	43.406 ms	3 - 216 MB	NPU
DDRNet23-Slim	TFLITE	float	Snapdragon® 8 Elite Mobile	15.633 ms	1 - 229 MB	NPU
DDRNet23-Slim	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	10.452 ms	2 - 244 MB	NPU

License

The license for the original implementation of DDRNet23-Slim 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/DDRNet23-Slim

Deep Dual-resolution Networks for Real-time and Accurate Semantic Segmentation of Road Scenes

Paper • 2101.06085 • Published Jan 15, 2021