MobileSam: Optimized for Qualcomm Devices

Transformer based encoder-decoder where prompts specify what to segment in an image thereby allowing segmentation without the need for additional training. The image encoder generates embeddings and the lightweight decoder operates on the embeddings for point and mask based image segmentation.

This is based on the implementation of MobileSam 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 2.42, ONNX Runtime 1.24.1	Download
QNN_DLC	float	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 MobileSam 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 MobileSam on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

• Model checkpoint: vit_t
• Input resolution: 720p (720x1280)
• Number of parameters (SAMEncoder): 6.95M
• Model size (SAMEncoder) (float): 26.6 MB
• Number of parameters (SAMDecoder): 6.16M
• Model size (SAMDecoder) (float): 23.7 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
MobileSAMDecoder	ONNX	float	Snapdragon® X Elite	7.515 ms	11 - 11 MB	NPU
MobileSAMDecoder	ONNX	float	Snapdragon® 8 Gen 3 Mobile	4.916 ms	5 - 263 MB	NPU
MobileSAMDecoder	ONNX	float	Qualcomm® QCS8550 (Proxy)	7.013 ms	0 - 15 MB	NPU
MobileSAMDecoder	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	3.65 ms	4 - 226 MB	NPU
MobileSAMDecoder	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	2.89 ms	0 - 232 MB	NPU
MobileSAMDecoder	ONNX	float	Snapdragon® X2 Elite	2.951 ms	11 - 11 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Snapdragon® X Elite	5.286 ms	4 - 4 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	3.36 ms	4 - 232 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	11.687 ms	1 - 198 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	4.86 ms	4 - 36 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Qualcomm® SA8775P	5.887 ms	0 - 200 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	8.484 ms	0 - 220 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Qualcomm® SA7255P	11.687 ms	1 - 198 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Qualcomm® SA8295P	7.197 ms	0 - 193 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	2.595 ms	0 - 198 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	2.154 ms	3 - 202 MB	NPU
MobileSAMDecoder	QNN_DLC	float	Snapdragon® X2 Elite	2.662 ms	4 - 4 MB	NPU
MobileSAMDecoder	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	3.875 ms	0 - 235 MB	NPU
MobileSAMDecoder	TFLITE	float	Qualcomm® QCS8275 (Proxy)	13.024 ms	0 - 203 MB	NPU
MobileSAMDecoder	TFLITE	float	Qualcomm® QCS8550 (Proxy)	5.6 ms	0 - 26 MB	NPU
MobileSAMDecoder	TFLITE	float	Qualcomm® SA8775P	6.7 ms	0 - 204 MB	NPU
MobileSAMDecoder	TFLITE	float	Qualcomm® QCS8450 (Proxy)	9.491 ms	0 - 225 MB	NPU
MobileSAMDecoder	TFLITE	float	Qualcomm® SA7255P	13.024 ms	0 - 203 MB	NPU
MobileSAMDecoder	TFLITE	float	Qualcomm® SA8295P	8.115 ms	0 - 201 MB	NPU
MobileSAMDecoder	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	2.955 ms	0 - 205 MB	NPU
MobileSAMDecoder	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	2.353 ms	0 - 206 MB	NPU
MobileSAMEncoder	ONNX	float	Snapdragon® X Elite	172.158 ms	12 - 12 MB	NPU
MobileSAMEncoder	ONNX	float	Snapdragon® 8 Gen 3 Mobile	125.387 ms	20 - 3042 MB	NPU
MobileSAMEncoder	ONNX	float	Qualcomm® QCS8550 (Proxy)	175.157 ms	12 - 31 MB	NPU
MobileSAMEncoder	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	102.261 ms	9 - 1390 MB	NPU
MobileSAMEncoder	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	83.472 ms	20 - 1524 MB	NPU
MobileSAMEncoder	ONNX	float	Snapdragon® X2 Elite	94.261 ms	14 - 14 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Snapdragon® X Elite	100.191 ms	12 - 12 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	66.146 ms	12 - 1636 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	209.193 ms	0 - 1209 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	96.518 ms	13 - 15 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Qualcomm® SA8775P	103.049 ms	0 - 1209 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	470.295 ms	13 - 1732 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Qualcomm® SA7255P	209.193 ms	0 - 1209 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Qualcomm® SA8295P	474.267 ms	0 - 1248 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	49.845 ms	12 - 1217 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	48.339 ms	10 - 1323 MB	NPU
MobileSAMEncoder	QNN_DLC	float	Snapdragon® X2 Elite	51.495 ms	12 - 12 MB	NPU
MobileSAMEncoder	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	127.145 ms	2 - 1509 MB	NPU
MobileSAMEncoder	TFLITE	float	Qualcomm® QCS8275 (Proxy)	295.631 ms	4 - 1269 MB	NPU
MobileSAMEncoder	TFLITE	float	Qualcomm® QCS8550 (Proxy)	175.109 ms	4 - 8 MB	NPU
MobileSAMEncoder	TFLITE	float	Qualcomm® SA8775P	180.289 ms	4 - 1269 MB	NPU
MobileSAMEncoder	TFLITE	float	Qualcomm® QCS8450 (Proxy)	584.211 ms	0 - 3174 MB	NPU
MobileSAMEncoder	TFLITE	float	Qualcomm® SA7255P	295.631 ms	4 - 1269 MB	NPU
MobileSAMEncoder	TFLITE	float	Qualcomm® SA8295P	502.945 ms	4 - 1893 MB	NPU
MobileSAMEncoder	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	95.905 ms	3 - 1261 MB	NPU
MobileSAMEncoder	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	78.087 ms	4 - 1349 MB	NPU

License

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

References

• Segment Anything
• Source Model Implementation

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.