SalsaNext: Optimized for Qualcomm Devices

SalsaNext is a LiDAR-based model designed for efficient and accurate semantic segmentation.

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 1.24.3	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

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

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

• Model checkpoint: SalsaNext
• Input resolution: 1x5x64x2048
• Number of parameters: 6.71M
• Model size (float): 25.7 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
SalsaNext	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	15.755 ms	25 - 220 MB	NPU
SalsaNext	ONNX	float	Snapdragon® 8 Elite Mobile	18.642 ms	23 - 211 MB	NPU
SalsaNext	ONNX	float	Snapdragon® X2 Elite	15.042 ms	33 - 33 MB	NPU
SalsaNext	ONNX	float	Snapdragon® X Elite	32.475 ms	33 - 33 MB	NPU
SalsaNext	ONNX	float	Snapdragon® X Elite	32.475 ms	33 - 33 MB	NPU
SalsaNext	ONNX	float	Snapdragon® 8 Gen 3 Mobile	25.465 ms	15 - 282 MB	NPU
SalsaNext	ONNX	float	Qualcomm® QCS8550 (Proxy)	34.192 ms	23 - 26 MB	NPU
SalsaNext	ONNX	float	Qualcomm® QCS9075	39.965 ms	23 - 26 MB	NPU
SalsaNext	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	18.642 ms	23 - 211 MB	NPU
SalsaNext	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	13.178 ms	3 - 290 MB	NPU
SalsaNext	QNN_DLC	float	Snapdragon® 8 Elite Mobile	18.513 ms	0 - 265 MB	NPU
SalsaNext	QNN_DLC	float	Snapdragon® X2 Elite	13.869 ms	3 - 3 MB	NPU
SalsaNext	QNN_DLC	float	Snapdragon® X Elite	31.941 ms	3 - 3 MB	NPU
SalsaNext	QNN_DLC	float	Snapdragon® X Elite	31.941 ms	3 - 3 MB	NPU
SalsaNext	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	25.515 ms	0 - 331 MB	NPU
SalsaNext	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	119.202 ms	1 - 247 MB	NPU
SalsaNext	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	34.069 ms	3 - 5 MB	NPU
SalsaNext	QNN_DLC	float	Qualcomm® QCS9075	37.589 ms	3 - 17 MB	NPU
SalsaNext	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	65.722 ms	3 - 348 MB	NPU
SalsaNext	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	18.513 ms	0 - 265 MB	NPU
SalsaNext	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	13.183 ms	9 - 295 MB	NPU
SalsaNext	TFLITE	float	Snapdragon® 8 Elite Mobile	18.589 ms	10 - 279 MB	NPU
SalsaNext	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	25.624 ms	10 - 347 MB	NPU
SalsaNext	TFLITE	float	Qualcomm® QCS8275 (Proxy)	118.708 ms	10 - 256 MB	NPU
SalsaNext	TFLITE	float	Qualcomm® QCS8550 (Proxy)	34.254 ms	10 - 12 MB	NPU
SalsaNext	TFLITE	float	Qualcomm® QCS9075	37.302 ms	10 - 39 MB	NPU
SalsaNext	TFLITE	float	Qualcomm® QCS8450 (Proxy)	65.293 ms	0 - 346 MB	NPU
SalsaNext	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	18.589 ms	10 - 279 MB	NPU

License

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

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.