Hugging Face's logo

qualcomm
/
Mask2Former

Image Segmentation
PyTorch
android
Model card Files
xet
Community
Mask2Former / README.md
qaihm-bot's picture
qaihm-bot
Upload README.md with huggingface_hub
9b69faa verified
preview code
|
raw
history blame
10.9 kB
metadata
library_name: pytorch
license: other
tags:
  - android
pipeline_tag: image-segmentation

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: Model_use_case.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 Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
Mask2Former float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 1886.595 ms 164 - 178 MB CPU Mask2Former.tflite
Mask2Former float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 1717.317 ms 164 - 190 MB CPU Mask2Former.tflite
Mask2Former float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 1444.853 ms 164 - 168 MB CPU Mask2Former.tflite
Mask2Former float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 1595.171 ms 164 - 179 MB CPU Mask2Former.tflite
Mask2Former float SA7255P ADP Qualcomm® SA7255P TFLITE 1886.595 ms 164 - 178 MB CPU Mask2Former.tflite
Mask2Former float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 1245.767 ms 163 - 188 MB CPU Mask2Former.tflite
Mask2Former float SA8295P ADP Qualcomm® SA8295P TFLITE 1397.371 ms 163 - 181 MB CPU Mask2Former.tflite
Mask2Former float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 1367.204 ms 138 - 169 MB CPU Mask2Former.tflite
Mask2Former float SA8775P ADP Qualcomm® SA8775P TFLITE 1595.171 ms 164 - 179 MB CPU Mask2Former.tflite
Mask2Former float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile TFLITE 1496.756 ms 152 - 171 MB CPU Mask2Former.tflite
Mask2Former float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile ONNX 958.304 ms 123 - 140 MB CPU Mask2Former.onnx
Mask2Former float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 1036.409 ms 106 - 134 MB CPU Mask2Former.tflite
Mask2Former float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 740.281 ms 232 - 258 MB CPU Mask2Former.onnx
Mask2Former float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile TFLITE 976.121 ms 69 - 89 MB CPU Mask2Former.tflite
Mask2Former float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile ONNX 689.35 ms 158 - 170 MB CPU Mask2Former.onnx
Mask2Former float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 513.005 ms 268 - 268 MB CPU Mask2Former.onnx

Installation

Install the package via pip:

pip install "qai-hub-models[mask2former]" git+https://github.com/cocodataset/panopticapi.git

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                          : cs_8275 (ANDROID 14)                  
Runtime                         : TFLITE                                
Estimated inference time (ms)   : 1886.6                                
Estimated peak memory usage (MB): [164, 178]                            
Total # Ops                     : 3213                                  
Compute Unit(s)                 : npu (0 ops) gpu (0 ops) cpu (3213 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