Mask2Former: Optimized for Mobile Deployment

Real-time object segmentation

Mask2Former is a machine learning model that predicts masks and classes of objects in an image.

This model is an implementation of Mask2Former found here.

This repository provides scripts to run Mask2Former on Qualcomm® devices. More details on model performance across various devices, can be found here.

Model Details

Model Type: Semantic segmentation
Model Stats:
- Model checkpoint: facebook/mask2former-swin-tiny-coco-panoptic
- Input resolution: 384x384
- Number of parameters: 42M
- Model size: 200.6 MB
- Number of output classes: 100

Model	Device	Chipset	Target Runtime	Inference Time (ms)	Peak Memory Range (MB)	Precision	Primary Compute Unit	Target Model
Mask2Former	Samsung Galaxy S23	Snapdragon® 8 Gen 2	TFLITE	1467.863 ms	164 - 169 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	Samsung Galaxy S23	Snapdragon® 8 Gen 2	ONNX	931.354 ms	131 - 142 MB	FP32	CPU	Mask2Former.onnx
Mask2Former	Samsung Galaxy S24	Snapdragon® 8 Gen 3	TFLITE	1229.494 ms	114 - 143 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	Samsung Galaxy S24	Snapdragon® 8 Gen 3	ONNX	874.774 ms	210 - 240 MB	FP32	CPU	Mask2Former.onnx
Mask2Former	Snapdragon 8 Elite QRD	Snapdragon® 8 Elite	TFLITE	967.366 ms	71 - 91 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	Snapdragon 8 Elite QRD	Snapdragon® 8 Elite	ONNX	685.02 ms	202 - 218 MB	FP32	CPU	Mask2Former.onnx
Mask2Former	SA7255P ADP	SA7255P	TFLITE	2251.173 ms	164 - 179 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	SA8255 (Proxy)	SA8255P Proxy	TFLITE	1530.724 ms	164 - 169 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	SA8295P ADP	SA8295P	TFLITE	1378.588 ms	163 - 185 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	SA8650 (Proxy)	SA8650P Proxy	TFLITE	1394.522 ms	157 - 162 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	SA8775P ADP	SA8775P	TFLITE	1837.887 ms	164 - 179 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	QCS8275 (Proxy)	QCS8275 Proxy	TFLITE	2251.173 ms	164 - 179 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	QCS8550 (Proxy)	QCS8550 Proxy	TFLITE	1332.317 ms	138 - 166 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	QCS9075 (Proxy)	QCS9075 Proxy	TFLITE	1837.887 ms	164 - 179 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	QCS8450 (Proxy)	QCS8450 Proxy	TFLITE	2028.103 ms	110 - 131 MB	FP32	CPU	Mask2Former.tflite
Mask2Former	Snapdragon X Elite CRD	Snapdragon® X Elite	ONNX	554.346 ms	267 - 267 MB	FP32	CPU	Mask2Former.onnx

Installation

Install the package via pip:

pip install "qai-hub-models[mask2former]"

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.mask2former.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.mask2former.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.mask2former.export

Profiling Results
------------------------------------------------------------
Mask2Former
Device                          : Samsung Galaxy S23 (13)
Runtime                         : TFLITE                 
Estimated inference time (ms)   : 1467.9                 
Estimated peak memory usage (MB): [164, 169]             
Total # Ops                     : 3223                   
Compute Unit(s)                 : CPU (3223 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.mask2former 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.mask2former.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.mask2former.demo -- --on-device

Deploying compiled model to Android

The models can be deployed using multiple runtimes:

TensorFlow Lite (.tflite export): This tutorial provides a guide to deploy the .tflite model in an Android application.
QNN (.so export ): This sample app provides instructions on how to use the .so shared library in an Android application.

View on Qualcomm® AI Hub

Get more details on Mask2Former's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

The license for the original implementation of Mask2Former can be found here.
The license for the compiled assets for on-device deployment can be found here

References

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.