Hugging Face's logo

qualcomm
/
Segformer-Base

Image Segmentation
PyTorch
LiteRT
ONNX
real_time
android
Model card Files Community

Segformer-Base: Optimized for Mobile Deployment

Real-time object segmentation

Segformer Base is a machine learning model that predicts masks and classes of objects in an image.

This model is an implementation of Segformer-Base found here.

This repository provides scripts to run Segformer-Base 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: nvidia/segformer-b0-finetuned-ade-512-512
    • Input resolution: 512x512
    • Number of parameters: 3.7M
    • Model size: 14.4 MB
    • Number of output classes: 150
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
Segformer-Base float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 315.099 ms 9 - 65 MB NPU Segformer-Base.tflite
Segformer-Base float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN 321.201 ms 2 - 11 MB NPU Use Export Script
Segformer-Base float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 114.238 ms 9 - 66 MB NPU Segformer-Base.tflite
Segformer-Base float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN 121.029 ms 3 - 55 MB NPU Use Export Script
Segformer-Base float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 115.018 ms 0 - 36 MB NPU Segformer-Base.tflite
Segformer-Base float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN 116.65 ms 3 - 6 MB NPU Use Export Script
Segformer-Base float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 117.193 ms 9 - 66 MB NPU Segformer-Base.tflite
Segformer-Base float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN 119.038 ms 1 - 10 MB NPU Use Export Script
Segformer-Base float SA7255P ADP Qualcomm® SA7255P TFLITE 315.099 ms 9 - 65 MB NPU Segformer-Base.tflite
Segformer-Base float SA7255P ADP Qualcomm® SA7255P QNN 321.201 ms 2 - 11 MB NPU Use Export Script
Segformer-Base float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 114.665 ms 0 - 38 MB NPU Segformer-Base.tflite
Segformer-Base float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN 116.561 ms 3 - 6 MB NPU Use Export Script
Segformer-Base float SA8295P ADP Qualcomm® SA8295P TFLITE 125.933 ms 9 - 61 MB NPU Segformer-Base.tflite
Segformer-Base float SA8295P ADP Qualcomm® SA8295P QNN 130.6 ms 0 - 17 MB NPU Use Export Script
Segformer-Base float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 114.784 ms 0 - 33 MB NPU Segformer-Base.tflite
Segformer-Base float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN 116.849 ms 3 - 5 MB NPU Use Export Script
Segformer-Base float SA8775P ADP Qualcomm® SA8775P TFLITE 117.193 ms 9 - 66 MB NPU Segformer-Base.tflite
Segformer-Base float SA8775P ADP Qualcomm® SA8775P QNN 119.038 ms 1 - 10 MB NPU Use Export Script
Segformer-Base float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile TFLITE 114.86 ms 0 - 36 MB NPU Segformer-Base.tflite
Segformer-Base float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile QNN 116.832 ms 3 - 18 MB NPU Use Export Script
Segformer-Base float Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile ONNX 111.55 ms 19 - 52 MB NPU Segformer-Base.onnx
Segformer-Base float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 86.521 ms 9 - 75 MB NPU Segformer-Base.tflite
Segformer-Base float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN 87.898 ms 3 - 62 MB NPU Use Export Script
Segformer-Base float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 83.731 ms 23 - 81 MB NPU Segformer-Base.onnx
Segformer-Base float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile TFLITE 93.336 ms 8 - 59 MB NPU Segformer-Base.tflite
Segformer-Base float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile QNN 97.56 ms 3 - 53 MB NPU Use Export Script
Segformer-Base float Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile ONNX 94.304 ms 23 - 70 MB NPU Segformer-Base.onnx
Segformer-Base float Snapdragon X Elite CRD Snapdragon® X Elite QNN 121.73 ms 3 - 3 MB NPU Use Export Script
Segformer-Base float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 118.315 ms 35 - 35 MB NPU Segformer-Base.onnx
Segformer-Base w8a16 Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile ONNX 40.96 ms 31 - 99 MB NPU Segformer-Base.onnx
Segformer-Base w8a16 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 28.528 ms 28 - 288 MB NPU Segformer-Base.onnx
Segformer-Base w8a16 Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile ONNX 31.159 ms 27 - 239 MB NPU Segformer-Base.onnx
Segformer-Base w8a16 Snapdragon X Elite CRD Snapdragon® X Elite ONNX 44.24 ms 58 - 58 MB NPU Segformer-Base.onnx
Segformer-Base w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 35.944 ms 2 - 41 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN 25.442 ms 1 - 10 MB NPU Use Export Script
Segformer-Base w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 13.76 ms 2 - 49 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN 9.382 ms 1 - 44 MB NPU Use Export Script
Segformer-Base w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 13.209 ms 2 - 14 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN 7.852 ms 1 - 4 MB NPU Use Export Script
Segformer-Base w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 13.755 ms 2 - 40 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN 8.462 ms 1 - 13 MB NPU Use Export Script
Segformer-Base w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) TFLITE 122.719 ms 14 - 55 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) QNN 33.447 ms 1 - 13 MB NPU Use Export Script
Segformer-Base w8a8 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) TFLITE 415.858 ms 50 - 68 MB GPU Segformer-Base.tflite
Segformer-Base w8a8 SA7255P ADP Qualcomm® SA7255P TFLITE 35.944 ms 2 - 41 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 SA7255P ADP Qualcomm® SA7255P QNN 25.442 ms 1 - 10 MB NPU Use Export Script
Segformer-Base w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 13.24 ms 2 - 15 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN 7.806 ms 1 - 4 MB NPU Use Export Script
Segformer-Base w8a8 SA8295P ADP Qualcomm® SA8295P TFLITE 15.501 ms 2 - 42 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 SA8295P ADP Qualcomm® SA8295P QNN 9.948 ms 1 - 18 MB NPU Use Export Script
Segformer-Base w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 13.25 ms 2 - 15 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN 7.81 ms 1 - 3 MB NPU Use Export Script
Segformer-Base w8a8 SA8775P ADP Qualcomm® SA8775P TFLITE 13.755 ms 2 - 40 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 SA8775P ADP Qualcomm® SA8775P QNN 8.462 ms 1 - 13 MB NPU Use Export Script
Segformer-Base w8a8 Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile TFLITE 13.244 ms 2 - 15 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile QNN 7.828 ms 1 - 12 MB NPU Use Export Script
Segformer-Base w8a8 Samsung Galaxy S23 Snapdragon® 8 Gen 2 Mobile ONNX 17.979 ms 3 - 25 MB NPU Segformer-Base.onnx
Segformer-Base w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 9.065 ms 2 - 52 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN 5.255 ms 1 - 61 MB NPU Use Export Script
Segformer-Base w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 12.47 ms 5 - 66 MB NPU Segformer-Base.onnx
Segformer-Base w8a8 Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile TFLITE 9.279 ms 2 - 43 MB NPU Segformer-Base.tflite
Segformer-Base w8a8 Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile QNN 4.788 ms 1 - 53 MB NPU Use Export Script
Segformer-Base w8a8 Snapdragon 8 Elite QRD Snapdragon® 8 Elite Mobile ONNX 11.956 ms 5 - 56 MB NPU Segformer-Base.onnx
Segformer-Base w8a8 Snapdragon X Elite CRD Snapdragon® X Elite QNN 8.692 ms 1 - 1 MB NPU Use Export Script
Segformer-Base w8a8 Snapdragon X Elite CRD Snapdragon® X Elite ONNX 19.372 ms 9 - 9 MB NPU Segformer-Base.onnx

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.segformer_base.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.segformer_base.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.segformer_base.export

Profiling Results
------------------------------------------------------------
Segformer-Base
Device                          : cs_8275 (ANDROID 14)                 
Runtime                         : TFLITE                               
Estimated inference time (ms)   : 315.1                                
Estimated peak memory usage (MB): [9, 65]                              
Total # Ops                     : 545                                  
Compute Unit(s)                 : npu (545 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.segformer_base 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.segformer_base.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.segformer_base.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 Segformer-Base's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

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

References

Community

Downloads last month
13
Inference Providers NEW
Image Segmentation
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support