library_name: pytorch
license: other
tags:
- android
pipeline_tag: depth-estimation
Midas-V2: Optimized for Mobile Deployment
Deep Convolutional Neural Network model for depth estimation
Midas is designed for estimating depth at each point in an image.
This model is an implementation of Midas-V2 found here.
This repository provides scripts to run Midas-V2 on Qualcomm® devices. More details on model performance across various devices, can be found here.
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): 16.6 MB
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model |
|---|---|---|---|---|---|---|---|---|
| Midas-V2 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 12.861 ms | 0 - 39 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 83.89 ms | 1 - 11 MB | NPU | Use Export Script |
| Midas-V2 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 4.967 ms | 0 - 49 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 7.389 ms | 0 - 36 MB | NPU | Use Export Script |
| Midas-V2 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 3.233 ms | 0 - 284 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 3.005 ms | 1 - 3 MB | NPU | Use Export Script |
| Midas-V2 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 4.544 ms | 0 - 40 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 4.237 ms | 1 - 15 MB | NPU | Use Export Script |
| Midas-V2 | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 12.861 ms | 0 - 39 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | SA7255P ADP | Qualcomm® SA7255P | QNN | 83.89 ms | 1 - 11 MB | NPU | Use Export Script |
| Midas-V2 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 3.24 ms | 0 - 272 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 3.01 ms | 1 - 3 MB | NPU | Use Export Script |
| Midas-V2 | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 5.746 ms | 0 - 25 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | SA8295P ADP | Qualcomm® SA8295P | QNN | 5.375 ms | 1 - 19 MB | NPU | Use Export Script |
| Midas-V2 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 3.241 ms | 0 - 247 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 3.011 ms | 0 - 2 MB | NPU | Use Export Script |
| Midas-V2 | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 4.544 ms | 0 - 40 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | SA8775P ADP | Qualcomm® SA8775P | QNN | 4.237 ms | 1 - 15 MB | NPU | Use Export Script |
| Midas-V2 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 3.23 ms | 0 - 248 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 3.034 ms | 0 - 16 MB | NPU | Use Export Script |
| Midas-V2 | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 2.961 ms | 0 - 73 MB | NPU | Midas-V2.onnx |
| Midas-V2 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 2.288 ms | 0 - 66 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 2.187 ms | 1 - 38 MB | NPU | Use Export Script |
| Midas-V2 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 2.068 ms | 0 - 44 MB | NPU | Midas-V2.onnx |
| Midas-V2 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 2.127 ms | 0 - 44 MB | NPU | Midas-V2.tflite |
| Midas-V2 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 1.909 ms | 1 - 30 MB | NPU | Use Export Script |
| Midas-V2 | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 1.92 ms | 1 - 31 MB | NPU | Midas-V2.onnx |
| Midas-V2 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 3.213 ms | 1 - 1 MB | NPU | Use Export Script |
| Midas-V2 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 3.068 ms | 36 - 36 MB | NPU | Midas-V2.onnx |
| Midas-V2 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 2.45 ms | 0 - 27 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 11.558 ms | 0 - 10 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 1.56 ms | 0 - 43 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 1.944 ms | 0 - 45 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 1.066 ms | 0 - 133 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 1.279 ms | 0 - 3 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 1.346 ms | 0 - 30 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 1.572 ms | 0 - 15 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | TFLITE | 3.774 ms | 0 - 44 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN | 5.582 ms | 0 - 15 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | TFLITE | 16.121 ms | 0 - 2 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 2.45 ms | 0 - 27 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN | 11.558 ms | 0 - 10 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 1.067 ms | 0 - 134 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN | 1.294 ms | 0 - 2 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 1.877 ms | 0 - 30 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN | 2.184 ms | 0 - 16 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 1.072 ms | 0 - 132 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN | 1.28 ms | 0 - 2 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 1.346 ms | 0 - 30 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN | 1.572 ms | 0 - 15 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 1.069 ms | 0 - 133 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 1.296 ms | 0 - 124 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 120.558 ms | 0 - 106 MB | NPU | Midas-V2.onnx |
| Midas-V2 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 0.763 ms | 0 - 57 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 0.905 ms | 0 - 55 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 93.185 ms | 16 - 349 MB | NPU | Midas-V2.onnx |
| Midas-V2 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 0.681 ms | 0 - 32 MB | NPU | Midas-V2.tflite |
| Midas-V2 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 0.786 ms | 0 - 34 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 82.77 ms | 25 - 346 MB | NPU | Midas-V2.onnx |
| Midas-V2 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 1.418 ms | 0 - 0 MB | NPU | Use Export Script |
| Midas-V2 | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 142.945 ms | 68 - 68 MB | NPU | Midas-V2.onnx |
Installation
Install the package via pip:
pip install "qai-hub-models[midas]"
Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
Sign-in to Qualcomm® AI Hub with your
Qualcomm® ID. Once signed in navigate to Account -> Settings -> API Token.
With this API token, you can configure your client to run models on the cloud hosted devices.
qai-hub configure --api_token API_TOKEN
Navigate to docs for more information.
Demo off target
The package contains a simple end-to-end demo that downloads pre-trained weights and runs this model on a sample input.
python -m qai_hub_models.models.midas.demo
The above demo runs a reference implementation of pre-processing, model inference, and post processing.
NOTE: If you want running in a Jupyter Notebook or Google Colab like environment, please add the following to your cell (instead of the above).
%run -m qai_hub_models.models.midas.demo
Run model on a cloud-hosted device
In addition to the demo, you can also run the model on a cloud-hosted Qualcomm® device. This script does the following:
- Performance check on-device on a cloud-hosted device
- Downloads compiled assets that can be deployed on-device for Android.
- Accuracy check between PyTorch and on-device outputs.
python -m qai_hub_models.models.midas.export
Profiling Results
------------------------------------------------------------
Midas-V2
Device : cs_8275 (ANDROID 14)
Runtime : TFLITE
Estimated inference time (ms) : 12.9
Estimated peak memory usage (MB): [0, 39]
Total # Ops : 138
Compute Unit(s) : npu (138 ops) gpu (0 ops) cpu (0 ops)
How does this work?
This export script leverages Qualcomm® AI Hub to optimize, validate, and deploy this model on-device. Lets go through each step below in detail:
Step 1: Compile model for on-device deployment
To compile a PyTorch model for on-device deployment, we first trace the model
in memory using the jit.trace and then call the submit_compile_job API.
import torch
import qai_hub as hub
from qai_hub_models.models.midas import Model
# Load the model
torch_model = Model.from_pretrained()
# Device
device = hub.Device("Samsung Galaxy S24")
# Trace model
input_shape = torch_model.get_input_spec()
sample_inputs = torch_model.sample_inputs()
pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
# Compile model on a specific device
compile_job = hub.submit_compile_job(
model=pt_model,
device=device,
input_specs=torch_model.get_input_spec(),
)
# Get target model to run on-device
target_model = compile_job.get_target_model()
Step 2: Performance profiling on cloud-hosted device
After compiling models from step 1. Models can be profiled model on-device using the
target_model. Note that this scripts runs the model on a device automatically
provisioned in the cloud. Once the job is submitted, you can navigate to a
provided job URL to view a variety of on-device performance metrics.
profile_job = hub.submit_profile_job(
model=target_model,
device=device,
)
Step 3: Verify on-device accuracy
To verify the accuracy of the model on-device, you can run on-device inference on sample input data on the same cloud hosted device.
input_data = torch_model.sample_inputs()
inference_job = hub.submit_inference_job(
model=target_model,
device=device,
inputs=input_data,
)
on_device_output = inference_job.download_output_data()
With the output of the model, you can compute like PSNR, relative errors or spot check the output with expected output.
Note: This on-device profiling and inference requires access to Qualcomm® AI Hub. Sign up for access.
Run demo on a cloud-hosted device
You can also run the demo on-device.
python -m qai_hub_models.models.midas.demo --on-device
NOTE: If you want running in a Jupyter Notebook or Google Colab like environment, please add the following to your cell (instead of the above).
%run -m qai_hub_models.models.midas.demo -- --on-device
Deploying compiled model to Android
The models can be deployed using multiple runtimes:
TensorFlow Lite (
.tfliteexport): This tutorial provides a guide to deploy the .tflite model in an Android application.QNN (
.soexport ): This sample app provides instructions on how to use the.soshared library in an Android application.
View on Qualcomm® AI Hub
Get more details on Midas-V2's performance across various devices here. Explore all available models on Qualcomm® AI Hub
License
- The license for the original implementation of Midas-V2 can be found here.
- The license for the compiled assets for on-device deployment 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.