--- library_name: pytorch license: other tags: - android pipeline_tag: image-segmentation --- ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/fastsam_s/web-assets/model_demo.png) # FastSam-S: Optimized for Mobile Deployment ## Generate high quality segmentation mask on device The Fast Segment Anything Model (FastSAM) is a novel, real-time CNN-based solution for the Segment Anything task. This task is designed to segment any object within an image based on various possible user interaction prompts. The model performs competitively despite significantly reduced computation, making it a practical choice for a variety of vision tasks. This model is an implementation of FastSam-S found [here](https://github.com/CASIA-IVA-Lab/FastSAM). This repository provides scripts to run FastSam-S on Qualcomm® devices. More details on model performance across various devices, can be found [here](https://aihub.qualcomm.com/models/fastsam_s). ### Model Details - **Model Type:** Model_use_case.semantic_segmentation - **Model Stats:** - Model checkpoint: fastsam-s.pt - Inference latency: RealTime - Input resolution: 640x640 - Number of parameters: 11.8M - Model size (float): 45.1 MB | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model |---|---|---|---|---|---|---|---|---| | FastSam-S | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 37.745 ms | 4 - 243 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 37.787 ms | 5 - 232 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 16.036 ms | 4 - 215 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 16.91 ms | 5 - 199 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 6.889 ms | 4 - 7 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 6.889 ms | 5 - 7 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 8.287 ms | 0 - 26 MB | NPU | [FastSam-S.onnx.zip](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx.zip) | | FastSam-S | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 10.782 ms | 4 - 232 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 10.761 ms | 1 - 233 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 37.745 ms | 4 - 243 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 37.787 ms | 5 - 232 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 12.913 ms | 4 - 179 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 12.842 ms | 0 - 162 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 10.782 ms | 4 - 232 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 10.761 ms | 1 - 233 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 5.165 ms | 4 - 402 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 5.183 ms | 5 - 382 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 6.031 ms | 16 - 219 MB | NPU | [FastSam-S.onnx.zip](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx.zip) | | FastSam-S | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 3.803 ms | 0 - 210 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 3.846 ms | 5 - 209 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 4.81 ms | 12 - 184 MB | NPU | [FastSam-S.onnx.zip](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx.zip) | | FastSam-S | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | TFLITE | 2.924 ms | 0 - 221 MB | NPU | [FastSam-S.tflite](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.tflite) | | FastSam-S | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | QNN_DLC | 2.97 ms | 5 - 203 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | ONNX | 3.6 ms | 2 - 155 MB | NPU | [FastSam-S.onnx.zip](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx.zip) | | FastSam-S | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 7.408 ms | 5 - 5 MB | NPU | [FastSam-S.dlc](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.dlc) | | FastSam-S | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 8.534 ms | 19 - 19 MB | NPU | [FastSam-S.onnx.zip](https://huggingface.co/qualcomm/FastSam-S/blob/main/FastSam-S.onnx.zip) | ## Installation Install the package via pip: ```bash # NOTE: 3.10 <= PYTHON_VERSION < 3.14 is supported. pip install "qai-hub-models[fastsam-s]" ``` ## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) 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. ```bash qai-hub configure --api_token API_TOKEN ``` Navigate to [docs](https://workbench.aihub.qualcomm.com/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. ```bash python -m qai_hub_models.models.fastsam_s.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.fastsam_s.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. ```bash python -m qai_hub_models.models.fastsam_s.export ``` ## How does this work? This [export script](https://aihub.qualcomm.com/models/fastsam_s/qai_hub_models/models/FastSam-S/export.py) leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) 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. ```python import torch import qai_hub as hub from qai_hub_models.models.fastsam_s 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. ```python 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. ```python 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](https://myaccount.qualcomm.com/signup). ## Run demo on a cloud-hosted device You can also run the demo on-device. ```bash python -m qai_hub_models.models.fastsam_s.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.fastsam_s.demo -- --eval-mode on-device ``` ## Deploying compiled model to Android The models can be deployed using multiple runtimes: - TensorFlow Lite (`.tflite` export): [This tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a guide to deploy the .tflite model in an Android application. - QNN (`.so` export ): This [sample app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html) provides instructions on how to use the `.so` shared library in an Android application. ## View on Qualcomm® AI Hub Get more details on FastSam-S's performance across various devices [here](https://aihub.qualcomm.com/models/fastsam_s). Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/) ## License * The license for the original implementation of FastSam-S can be found [here](https://github.com/CASIA-IVA-Lab/FastSAM/blob/main/LICENSE). ## References * [Fast Segment Anything](https://arxiv.org/abs/2306.12156) * [Source Model Implementation](https://github.com/CASIA-IVA-Lab/FastSAM) ## Community * Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI. * For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).