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 1.24.1 | Download |
| ONNX | w8a16 | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | Download |
| QNN_DLC | float | Universal | QAIRT 2.43 | Download |
| QNN_DLC | w8a16 | Universal | QAIRT 2.43 | Download |
| TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | 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® X Elite | 32.445 ms | 33 - 33 MB | NPU |
| SalsaNext | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 25.95 ms | 24 - 293 MB | NPU |
| SalsaNext | ONNX | float | Qualcomm® QCS8550 (Proxy) | 33.816 ms | 23 - 27 MB | NPU |
| SalsaNext | ONNX | float | Qualcomm® QCS9075 | 40.161 ms | 23 - 25 MB | NPU |
| SalsaNext | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 18.578 ms | 24 - 208 MB | NPU |
| SalsaNext | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 15.642 ms | 14 - 205 MB | NPU |
| SalsaNext | ONNX | float | Snapdragon® X2 Elite | 14.991 ms | 33 - 33 MB | NPU |
| SalsaNext | ONNX | w8a16 | Snapdragon® X Elite | 34.386 ms | 18 - 18 MB | NPU |
| SalsaNext | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 29.077 ms | 16 - 304 MB | NPU |
| SalsaNext | ONNX | w8a16 | Qualcomm® QCS6490 | 3795.948 ms | 357 - 361 MB | CPU |
| SalsaNext | ONNX | w8a16 | Qualcomm® QCS8550 (Proxy) | 36.354 ms | 10 - 22 MB | NPU |
| SalsaNext | ONNX | w8a16 | Qualcomm® QCS9075 | 34.25 ms | 15 - 17 MB | NPU |
| SalsaNext | ONNX | w8a16 | Qualcomm® QCM6690 | 2029.889 ms | 291 - 299 MB | CPU |
| SalsaNext | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 23.898 ms | 13 - 211 MB | NPU |
| SalsaNext | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 1831.517 ms | 367 - 376 MB | CPU |
| SalsaNext | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 18.521 ms | 13 - 209 MB | NPU |
| SalsaNext | ONNX | w8a16 | Snapdragon® X2 Elite | 16.548 ms | 16 - 16 MB | NPU |
| SalsaNext | QNN_DLC | float | Snapdragon® X Elite | 30.082 ms | 3 - 3 MB | NPU |
| SalsaNext | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 22.329 ms | 3 - 267 MB | NPU |
| SalsaNext | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 116.597 ms | 1 - 177 MB | NPU |
| SalsaNext | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 30.73 ms | 3 - 4 MB | NPU |
| SalsaNext | QNN_DLC | float | Qualcomm® QCS9075 | 40.157 ms | 3 - 17 MB | NPU |
| SalsaNext | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 55.247 ms | 2 - 298 MB | NPU |
| SalsaNext | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 21.152 ms | 3 - 187 MB | NPU |
| SalsaNext | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 15.35 ms | 3 - 190 MB | NPU |
| SalsaNext | QNN_DLC | float | Snapdragon® X2 Elite | 15.175 ms | 3 - 3 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Snapdragon® X Elite | 43.777 ms | 1 - 1 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Snapdragon® 8 Gen 3 Mobile | 34.672 ms | 1 - 366 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Qualcomm® QCS8275 (Proxy) | 76.754 ms | 1 - 257 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Qualcomm® QCS8550 (Proxy) | 45.011 ms | 1 - 3 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Qualcomm® QCS9075 | 50.536 ms | 0 - 8 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Qualcomm® QCS8450 (Proxy) | 83.039 ms | 1 - 423 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 28.341 ms | 1 - 275 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 19.89 ms | 1 - 314 MB | NPU |
| SalsaNext | QNN_DLC | w8a16 | Snapdragon® X2 Elite | 18.42 ms | 1 - 1 MB | NPU |
| SalsaNext | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 22.506 ms | 0 - 269 MB | NPU |
| SalsaNext | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 116.821 ms | 10 - 188 MB | NPU |
| SalsaNext | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 30.418 ms | 10 - 26 MB | NPU |
| SalsaNext | TFLITE | float | Qualcomm® QCS9075 | 39.655 ms | 10 - 38 MB | NPU |
| SalsaNext | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 55.361 ms | 10 - 310 MB | NPU |
| SalsaNext | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 18.394 ms | 8 - 195 MB | NPU |
| SalsaNext | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 15.304 ms | 10 - 199 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.