Midas-V2: Optimized for Qualcomm Devices

Midas is designed for estimating depth at each point in an image.

This is based on the implementation of Midas-V2 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
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 Midas-V2 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 Midas-V2 on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.depth_estimation

Model Stats:

• Model checkpoint: MiDaS_small
• Input resolution: 256x256
• Number of parameters: 16.6M
• Model size (float): 63.2 MB
• Model size (w8a8): 16.9 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Midas-V2	ONNX	float	Snapdragon® X Elite	2.893 ms	36 - 36 MB	NPU
Midas-V2	ONNX	float	Snapdragon® 8 Gen 3 Mobile	2.117 ms	0 - 150 MB	NPU
Midas-V2	ONNX	float	Qualcomm® QCS8550 (Proxy)	3.065 ms	0 - 107 MB	NPU
Midas-V2	ONNX	float	Qualcomm® QCS9075	4.251 ms	1 - 4 MB	NPU
Midas-V2	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	1.691 ms	0 - 111 MB	NPU
Midas-V2	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	1.4 ms	0 - 111 MB	NPU
Midas-V2	QNN_DLC	float	Snapdragon® X Elite	3.308 ms	1 - 1 MB	NPU
Midas-V2	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	2.173 ms	1 - 199 MB	NPU
Midas-V2	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	12.08 ms	1 - 161 MB	NPU
Midas-V2	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	3.074 ms	1 - 3 MB	NPU
Midas-V2	QNN_DLC	float	Qualcomm® SA8775P	4.29 ms	1 - 161 MB	NPU
Midas-V2	QNN_DLC	float	Qualcomm® QCS9075	4.497 ms	1 - 3 MB	NPU
Midas-V2	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	6.801 ms	0 - 192 MB	NPU
Midas-V2	QNN_DLC	float	Qualcomm® SA7255P	12.08 ms	1 - 161 MB	NPU
Midas-V2	QNN_DLC	float	Qualcomm® SA8295P	5.452 ms	1 - 164 MB	NPU
Midas-V2	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	1.62 ms	0 - 162 MB	NPU
Midas-V2	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	1.287 ms	1 - 166 MB	NPU
Midas-V2	QNN_DLC	w8a8	Snapdragon® X Elite	1.479 ms	0 - 0 MB	NPU
Midas-V2	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	0.917 ms	0 - 80 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® QCS6490	4.171 ms	0 - 2 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	2.907 ms	0 - 46 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	1.32 ms	0 - 2 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® SA8775P	1.576 ms	0 - 50 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® QCS9075	1.439 ms	2 - 4 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® QCM6690	9.029 ms	0 - 174 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	1.833 ms	0 - 80 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® SA7255P	2.907 ms	0 - 46 MB	NPU
Midas-V2	QNN_DLC	w8a8	Qualcomm® SA8295P	2.237 ms	0 - 45 MB	NPU
Midas-V2	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.681 ms	0 - 47 MB	NPU
Midas-V2	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	1.585 ms	0 - 169 MB	NPU
Midas-V2	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.556 ms	0 - 50 MB	NPU
Midas-V2	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	2.151 ms	0 - 231 MB	NPU
Midas-V2	TFLITE	float	Qualcomm® QCS8275 (Proxy)	12.113 ms	0 - 182 MB	NPU
Midas-V2	TFLITE	float	Qualcomm® QCS8550 (Proxy)	3.068 ms	0 - 2 MB	NPU
Midas-V2	TFLITE	float	Qualcomm® SA8775P	4.328 ms	0 - 181 MB	NPU
Midas-V2	TFLITE	float	Qualcomm® QCS9075	4.441 ms	0 - 39 MB	NPU
Midas-V2	TFLITE	float	Qualcomm® QCS8450 (Proxy)	6.849 ms	0 - 215 MB	NPU
Midas-V2	TFLITE	float	Qualcomm® SA7255P	12.113 ms	0 - 182 MB	NPU
Midas-V2	TFLITE	float	Qualcomm® SA8295P	5.46 ms	0 - 165 MB	NPU
Midas-V2	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	1.619 ms	0 - 178 MB	NPU
Midas-V2	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	1.287 ms	0 - 179 MB	NPU
Midas-V2	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	0.749 ms	0 - 79 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® QCS6490	3.66 ms	0 - 28 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	2.494 ms	0 - 46 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	1.065 ms	0 - 4 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® SA8775P	1.364 ms	0 - 49 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® QCS9075	1.191 ms	0 - 20 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® QCM6690	8.394 ms	0 - 171 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	1.568 ms	0 - 77 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® SA7255P	2.494 ms	0 - 46 MB	NPU
Midas-V2	TFLITE	w8a8	Qualcomm® SA8295P	1.944 ms	0 - 43 MB	NPU
Midas-V2	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	0.597 ms	0 - 52 MB	NPU
Midas-V2	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	1.343 ms	0 - 169 MB	NPU
Midas-V2	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	0.49 ms	0 - 49 MB	NPU

License

• The license for the original implementation of Midas-V2 can be found here.

References

• Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer
• 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.