Hugging Face's logo

qualcomm
/
Segment-Anything-Model-2

Image Segmentation
PyTorch
LiteRT
foundation
android
Model card Files
xet
Community

Segment-Anything-Model-2: Optimized for Mobile Deployment

High-quality segmentation in images and videos with real-time performance and minimal user interaction

SAM 2, the successor to Meta's Segment Anything Model (SAM), is a cutting-edge tool designed for comprehensive object segmentation in both images and videos. It excels in handling complex visual data through a unified, promptable model architecture that supports real-time processing and zero-shot generalization.

This model is an implementation of Segment-Anything-Model-2 found here.

This repository provides scripts to run Segment-Anything-Model-2 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: sam2.1_hiera_t
    • Input resolution: 720p (720x1280)
    • Number of parameters (SAM2Encoder): 33.5M
    • Model size (SAM2Encoder) (float): 128 MB
    • Number of parameters (SAM2Decoder): 6.22M
    • Model size (SAM2Decoder) (float): 23.7 MB
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
SAM2Encoder float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 571.663 ms 16 - 1186 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 468.455 ms 414 - 1262 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 345.016 ms 16 - 2453 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 390.43 ms 12 - 2690 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 262.04 ms 16 - 99 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 214.433 ms 12 - 91 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 274.526 ms 16 - 1185 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 226.111 ms 0 - 847 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float SA7255P ADP Qualcomm® SA7255P TFLITE 571.663 ms 16 - 1186 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float SA7255P ADP Qualcomm® SA7255P QNN_DLC 468.455 ms 414 - 1262 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 262.304 ms 16 - 95 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 217.939 ms 12 - 92 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float SA8295P ADP Qualcomm® SA8295P TFLITE 378.186 ms 10 - 1141 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 261.486 ms 16 - 88 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 215.116 ms 12 - 90 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float SA8775P ADP Qualcomm® SA8775P TFLITE 274.526 ms 16 - 1185 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float SA8775P ADP Qualcomm® SA8775P QNN_DLC 226.111 ms 0 - 847 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 191.454 ms 14 - 2402 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 157.619 ms 12 - 866 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 155.037 ms 15 - 1267 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 125.904 ms 12 - 836 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 102.257 ms 12 - 832 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder float Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 204.535 ms 166 - 166 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 14.21 ms 0 - 48 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 12.322 ms 4 - 57 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 8.436 ms 0 - 53 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 13.594 ms 16 - 68 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 5.92 ms 0 - 24 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 5.239 ms 16 - 38 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 7.241 ms 0 - 48 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 6.655 ms 0 - 55 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float SA7255P ADP Qualcomm® SA7255P TFLITE 14.21 ms 0 - 48 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float SA7255P ADP Qualcomm® SA7255P QNN_DLC 12.322 ms 4 - 57 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 5.921 ms 0 - 30 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 5.307 ms 16 - 31 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float SA8295P ADP Qualcomm® SA8295P TFLITE 9.56 ms 0 - 51 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 5.884 ms 0 - 27 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 5.296 ms 16 - 29 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float SA8775P ADP Qualcomm® SA8775P TFLITE 7.241 ms 0 - 48 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float SA8775P ADP Qualcomm® SA8775P QNN_DLC 6.655 ms 0 - 55 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 4.138 ms 0 - 56 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 3.627 ms 0 - 66 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 3.546 ms 0 - 53 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 2.918 ms 0 - 67 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 2.534 ms 4 - 54 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder float Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 5.47 ms 16 - 16 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 226.387 ms 4 - 881 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 152.642 ms 3 - 1056 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 140.593 ms 4 - 1097 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 156.383 ms 3 - 1304 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 131.105 ms 0 - 115 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 93.753 ms 0 - 123 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 132.467 ms 4 - 881 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 95.728 ms 3 - 1059 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) TFLITE 1716.135 ms 71 - 307 MB CPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) QNN_DLC 817.975 ms 3 - 1514 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 SA7255P ADP Qualcomm® SA7255P TFLITE 226.387 ms 4 - 881 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 SA7255P ADP Qualcomm® SA7255P QNN_DLC 152.642 ms 3 - 1056 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 130.383 ms 0 - 115 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 96.789 ms 0 - 128 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 SA8295P ADP Qualcomm® SA8295P TFLITE 152.717 ms 4 - 887 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 SA8295P ADP Qualcomm® SA8295P QNN_DLC 108.5 ms 3 - 1056 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 130.887 ms 0 - 114 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 96.484 ms 0 - 123 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 SA8775P ADP Qualcomm® SA8775P TFLITE 132.467 ms 4 - 881 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 SA8775P ADP Qualcomm® SA8775P QNN_DLC 95.728 ms 3 - 1059 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 93.563 ms 4 - 963 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 67.284 ms 3 - 1162 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 73.158 ms 2 - 888 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 51.362 ms 3 - 1029 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 Snapdragon 7 Gen 4 QRD Snapdragon® 7 Gen 4 Mobile TFLITE 1232.584 ms 105 - 334 MB CPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 Snapdragon 7 Gen 4 QRD Snapdragon® 7 Gen 4 Mobile QNN_DLC 129.176 ms 3 - 1632 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile TFLITE 72.485 ms 3 - 924 MB NPU Segment-Anything-Model-2.tflite
SAM2Encoder w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 45.59 ms 3 - 1576 MB NPU Segment-Anything-Model-2.dlc
SAM2Encoder w8a8 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 96.742 ms 277 - 277 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 21.767 ms 0 - 64 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 3.874 ms 4 - 56 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 13.575 ms 0 - 71 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 2.685 ms 4 - 67 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 12.941 ms 0 - 49 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 1.925 ms 4 - 17 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 51.253 ms 0 - 64 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 2.587 ms 4 - 55 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) TFLITE 49.213 ms 13 - 66 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) QNN_DLC 12.293 ms 4 - 72 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 SA7255P ADP Qualcomm® SA7255P TFLITE 21.767 ms 0 - 64 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 SA7255P ADP Qualcomm® SA7255P QNN_DLC 3.874 ms 4 - 56 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 12.909 ms 0 - 50 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 1.923 ms 4 - 16 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 SA8295P ADP Qualcomm® SA8295P TFLITE 14.849 ms 0 - 65 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 SA8295P ADP Qualcomm® SA8295P QNN_DLC 2.895 ms 4 - 62 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 12.859 ms 0 - 49 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 1.922 ms 4 - 18 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 SA8775P ADP Qualcomm® SA8775P TFLITE 51.253 ms 0 - 64 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 SA8775P ADP Qualcomm® SA8775P QNN_DLC 2.587 ms 4 - 55 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 9.743 ms 0 - 72 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 1.258 ms 1 - 66 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 7.73 ms 0 - 69 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 0.997 ms 1 - 58 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 Snapdragon 7 Gen 4 QRD Snapdragon® 7 Gen 4 Mobile TFLITE 16.845 ms 12 - 54 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 Snapdragon 7 Gen 4 QRD Snapdragon® 7 Gen 4 Mobile QNN_DLC 2.715 ms 4 - 62 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile TFLITE 7.78 ms 0 - 67 MB NPU Segment-Anything-Model-2.tflite
SAM2Decoder w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 0.895 ms 1 - 63 MB NPU Segment-Anything-Model-2.dlc
SAM2Decoder w8a8 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 2.265 ms 23 - 23 MB NPU Segment-Anything-Model-2.dlc

Installation

Install the package via pip:

# NOTE: 3.10 <= PYTHON_VERSION < 3.14 is supported.
pip install "qai-hub-models[sam2]" git+https://github.com/facebookresearch/sam2.git@2b90b9f

Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device

Sign-in to Qualcomm® AI Hub Workbench 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.sam2.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.sam2.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.sam2.export

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.sam2 import Model

# Load the model
torch_model = Model.from_pretrained()

# Device
device = hub.Device("Samsung Galaxy S25")

# 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 Workbench. 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.sam2.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.sam2.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 Segment-Anything-Model-2's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

  • The license for the original implementation of Segment-Anything-Model-2 can be found here.
  • The license for the compiled assets for on-device deployment can be found here

References

Community

Downloads last month
141
Inference Providers NEW
Image Segmentation
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support