Hugging Face's logo

qualcomm
/
FCN-ResNet50

Image Segmentation
PyTorch
LiteRT
android
Model card Files Community
FCN-ResNet50 / README.md
qaihm-bot's picture
qaihm-bot
v0.35.0
2870421 verified
preview code
|
raw
history blame contribute delete
19 kB
metadata
library_name: pytorch
license: other
tags:
  - android
pipeline_tag: image-segmentation

FCN-ResNet50: Optimized for Mobile Deployment

Fully-convolutional network model for image segmentation

FCN_ResNet50 is a machine learning model that can segment images from the COCO dataset. It uses ResNet50 as a backbone.

This model is an implementation of FCN-ResNet50 found here.

This repository provides scripts to run FCN-ResNet50 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: COCO_WITH_VOC_LABELS_V1
    • Input resolution: 224x224
    • Number of output classes: 21
    • Number of parameters: 33.0M
    • Model size (float): 126 MB
    • Model size (w8a8): 32.2 MB
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
FCN-ResNet50 float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 277.436 ms 0 - 133 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 286.762 ms 0 - 188 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 74.583 ms 0 - 121 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 100.538 ms 3 - 97 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 47.438 ms 0 - 22 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 59.347 ms 3 - 65 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 77.225 ms 0 - 133 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 86.085 ms 3 - 193 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float SA7255P ADP Qualcomm® SA7255P TFLITE 277.436 ms 0 - 133 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float SA7255P ADP Qualcomm® SA7255P QNN_DLC 286.762 ms 0 - 188 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 47.227 ms 0 - 22 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 59.24 ms 3 - 67 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float SA8295P ADP Qualcomm® SA8295P TFLITE 84.449 ms 0 - 96 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float SA8295P ADP Qualcomm® SA8295P QNN_DLC 93.247 ms 0 - 97 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 47.578 ms 0 - 21 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 58.383 ms 3 - 58 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float SA8775P ADP Qualcomm® SA8775P TFLITE 77.225 ms 0 - 133 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float SA8775P ADP Qualcomm® SA8775P QNN_DLC 86.085 ms 3 - 193 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile TFLITE 46.987 ms 0 - 22 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile QNN_DLC 59.032 ms 3 - 63 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile ONNX 42.728 ms 0 - 109 MB NPU FCN-ResNet50.onnx.zip
FCN-ResNet50 float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 35.202 ms 0 - 157 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 45.845 ms 3 - 200 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 32.513 ms 3 - 121 MB NPU FCN-ResNet50.onnx.zip
FCN-ResNet50 float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile TFLITE 34.299 ms 0 - 135 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile QNN_DLC 42.928 ms 3 - 213 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile ONNX 31.304 ms 3 - 108 MB NPU FCN-ResNet50.onnx.zip
FCN-ResNet50 float Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 56.355 ms 75 - 75 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 44.241 ms 63 - 63 MB NPU FCN-ResNet50.onnx.zip
FCN-ResNet50 w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 39.449 ms 0 - 52 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 39.025 ms 1 - 78 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 17.366 ms 0 - 91 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 26.868 ms 1 - 97 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 14.603 ms 0 - 11 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 15.06 ms 1 - 18 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 15.034 ms 0 - 52 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 15.086 ms 1 - 66 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) TFLITE 110.77 ms 0 - 130 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) QNN_DLC 137.872 ms 0 - 159 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) TFLITE 1386.044 ms 69 - 146 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 SA7255P ADP Qualcomm® SA7255P TFLITE 39.449 ms 0 - 52 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 SA7255P ADP Qualcomm® SA7255P QNN_DLC 39.025 ms 1 - 78 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 14.605 ms 0 - 109 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 15.043 ms 3 - 21 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 SA8295P ADP Qualcomm® SA8295P TFLITE 21.797 ms 0 - 58 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 SA8295P ADP Qualcomm® SA8295P QNN_DLC 21.578 ms 1 - 78 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 14.643 ms 0 - 16 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 14.934 ms 0 - 20 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 SA8775P ADP Qualcomm® SA8775P TFLITE 15.034 ms 0 - 52 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 SA8775P ADP Qualcomm® SA8775P QNN_DLC 15.086 ms 1 - 66 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile TFLITE 14.601 ms 0 - 18 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile QNN_DLC 15.033 ms 0 - 19 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 10.967 ms 0 - 82 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 11.685 ms 1 - 108 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile TFLITE 10.673 ms 0 - 57 MB NPU FCN-ResNet50.tflite
FCN-ResNet50 w8a8 Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile QNN_DLC 10.021 ms 1 - 71 MB NPU FCN-ResNet50.dlc
FCN-ResNet50 w8a8 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 15.09 ms 104 - 104 MB NPU FCN-ResNet50.dlc

Installation

Install the package via pip:

pip install qai-hub-models

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.fcn_resnet50.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.fcn_resnet50.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.fcn_resnet50.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.fcn_resnet50 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.fcn_resnet50.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.fcn_resnet50.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 FCN-ResNet50's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

  • The license for the original implementation of FCN-ResNet50 can be found here.
  • The license for the compiled assets for on-device deployment can be found here

References

Community