Hugging Face's logo

qualcomm
/
MobileSam

Image Segmentation
PyTorch
LiteRT
foundation
android
Model card Files Community

MobileSam: Optimized for Mobile Deployment

Faster Segment Anything: Towards lightweight SAM for mobile applications

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 model is an implementation of MobileSam found here.

This repository provides scripts to run MobileSam 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: 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
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
MobileSAMEncoder float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 744.282 ms 33 - 379 MB NPU MobileSam.tflite
MobileSAMEncoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 592.073 ms 33 - 729 MB NPU MobileSam.tflite
MobileSAMEncoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 490.954 ms 12 - 618 MB NPU MobileSam.dlc
MobileSAMEncoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 438.155 ms 33 - 62 MB NPU MobileSam.tflite
MobileSAMEncoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 271.374 ms 12 - 86 MB NPU MobileSam.dlc
MobileSAMEncoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 424.391 ms 33 - 376 MB NPU MobileSam.tflite
MobileSAMEncoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 277.597 ms 1 - 1161 MB NPU MobileSam.dlc
MobileSAMEncoder float SA7255P ADP Qualcomm® SA7255P TFLITE 744.282 ms 33 - 379 MB NPU MobileSam.tflite
MobileSAMEncoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 425.325 ms 33 - 56 MB NPU MobileSam.tflite
MobileSAMEncoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 273.237 ms 12 - 85 MB NPU MobileSam.dlc
MobileSAMEncoder float SA8295P ADP Qualcomm® SA8295P TFLITE 584.827 ms 31 - 378 MB NPU MobileSam.tflite
MobileSAMEncoder float SA8295P ADP Qualcomm® SA8295P QNN_DLC 436.56 ms 3 - 650 MB NPU MobileSam.dlc
MobileSAMEncoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 438.208 ms 26 - 58 MB NPU MobileSam.tflite
MobileSAMEncoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 275.045 ms 12 - 87 MB NPU MobileSam.dlc
MobileSAMEncoder float SA8775P ADP Qualcomm® SA8775P TFLITE 424.391 ms 33 - 376 MB NPU MobileSam.tflite
MobileSAMEncoder float SA8775P ADP Qualcomm® SA8775P QNN_DLC 277.597 ms 1 - 1161 MB NPU MobileSam.dlc
MobileSAMEncoder float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile TFLITE 460.958 ms 33 - 58 MB NPU MobileSam.tflite
MobileSAMEncoder float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile QNN_DLC 271.764 ms 12 - 73 MB NPU MobileSam.dlc
MobileSAMEncoder float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile ONNX 413.682 ms 80 - 174 MB NPU MobileSam.onnx
MobileSAMEncoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 348.17 ms 33 - 721 MB NPU MobileSam.tflite
MobileSAMEncoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 205.547 ms 12 - 2423 MB NPU MobileSam.dlc
MobileSAMEncoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 312.791 ms 110 - 832 MB NPU MobileSam.onnx
MobileSAMEncoder float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile TFLITE 273.427 ms 33 - 379 MB NPU MobileSam.tflite
MobileSAMEncoder float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile QNN_DLC 173.722 ms 8 - 1176 MB NPU MobileSam.dlc
MobileSAMEncoder float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile ONNX 266.959 ms 128 - 531 MB NPU MobileSam.onnx
MobileSAMEncoder float Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 297.893 ms 48 - 48 MB NPU MobileSam.dlc
MobileSAMEncoder float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 466.886 ms 130 - 130 MB NPU MobileSam.onnx
MobileSAMDecoder float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 16.575 ms 0 - 54 MB NPU MobileSam.tflite
MobileSAMDecoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 8.251 ms 0 - 57 MB NPU MobileSam.tflite
MobileSAMDecoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 8.386 ms 4 - 54 MB NPU MobileSam.dlc
MobileSAMDecoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 7.302 ms 0 - 29 MB NPU MobileSam.tflite
MobileSAMDecoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 6.097 ms 4 - 21 MB NPU MobileSam.dlc
MobileSAMDecoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 8.429 ms 0 - 55 MB NPU MobileSam.tflite
MobileSAMDecoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 7.328 ms 0 - 45 MB NPU MobileSam.dlc
MobileSAMDecoder float SA7255P ADP Qualcomm® SA7255P TFLITE 16.575 ms 0 - 54 MB NPU MobileSam.tflite
MobileSAMDecoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 7.262 ms 0 - 28 MB NPU MobileSam.tflite
MobileSAMDecoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 6.112 ms 4 - 19 MB NPU MobileSam.dlc
MobileSAMDecoder float SA8295P ADP Qualcomm® SA8295P TFLITE 9.685 ms 0 - 50 MB NPU MobileSam.tflite
MobileSAMDecoder float SA8295P ADP Qualcomm® SA8295P QNN_DLC 7.457 ms 1 - 62 MB NPU MobileSam.dlc
MobileSAMDecoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 7.272 ms 0 - 32 MB NPU MobileSam.tflite
MobileSAMDecoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 6.109 ms 4 - 24 MB NPU MobileSam.dlc
MobileSAMDecoder float SA8775P ADP Qualcomm® SA8775P TFLITE 8.429 ms 0 - 55 MB NPU MobileSam.tflite
MobileSAMDecoder float SA8775P ADP Qualcomm® SA8775P QNN_DLC 7.328 ms 0 - 45 MB NPU MobileSam.dlc
MobileSAMDecoder float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile TFLITE 7.312 ms 0 - 29 MB NPU MobileSam.tflite
MobileSAMDecoder float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile QNN_DLC 6.103 ms 3 - 27 MB NPU MobileSam.dlc
MobileSAMDecoder float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile ONNX 9.413 ms 0 - 47 MB NPU MobileSam.onnx
MobileSAMDecoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 5.054 ms 0 - 62 MB NPU MobileSam.tflite
MobileSAMDecoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 4.198 ms 4 - 59 MB NPU MobileSam.dlc
MobileSAMDecoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 6.161 ms 4 - 72 MB NPU MobileSam.onnx
MobileSAMDecoder float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile TFLITE 5.065 ms 0 - 58 MB NPU MobileSam.tflite
MobileSAMDecoder float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile QNN_DLC 3.093 ms 4 - 54 MB NPU MobileSam.dlc
MobileSAMDecoder float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile ONNX 5.454 ms 1 - 60 MB NPU MobileSam.onnx
MobileSAMDecoder float Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 8.001 ms 12 - 12 MB NPU MobileSam.dlc
MobileSAMDecoder float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 10.193 ms 12 - 12 MB NPU MobileSam.onnx

Installation

Install the package via pip:

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

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.mobilesam.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.mobilesam.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.mobilesam.export

Profiling Results
------------------------------------------------------------
MobileSAMEncoder
Device                          : cs_8275 (ANDROID 14)                  
Runtime                         : TFLITE                                
Estimated inference time (ms)   : 744.3                                 
Estimated peak memory usage (MB): [33, 379]                             
Total # Ops                     : 592                                   
Compute Unit(s)                 : npu (532 ops) gpu (0 ops) cpu (60 ops)

------------------------------------------------------------
MobileSAMDecoder
Device                          : cs_8275 (ANDROID 14)                 
Runtime                         : TFLITE                               
Estimated inference time (ms)   : 16.6                                 
Estimated peak memory usage (MB): [0, 54]                              
Total # Ops                     : 846                                  
Compute Unit(s)                 : npu (846 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.mobilesam 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.mobilesam.demo --eval-mode 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.mobilesam.demo -- --eval-mode 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 MobileSam's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

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

References

Community

Downloads last month
53
Inference Providers NEW
Image Segmentation
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support