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 output classes: 150
- Number of parameters: 3.75M
- Model size (float): 14.4 MB
- Model size (w8a16): 4.57 MB
- Model size (w8a8): 3.90 MB
| 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 | 218.091 ms | 7 - 56 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 214.118 ms | 3 - 47 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 112.468 ms | 9 - 68 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 111.857 ms | 3 - 59 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 111.553 ms | 9 - 25 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 107.503 ms | 3 - 19 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 114.381 ms | 9 - 58 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 110.089 ms | 3 - 49 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 218.091 ms | 7 - 56 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 214.118 ms | 3 - 47 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 111.709 ms | 9 - 18 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 107.559 ms | 4 - 25 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 124.241 ms | 10 - 59 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 121.413 ms | 2 - 50 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 111.886 ms | 9 - 22 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 108.177 ms | 2 - 18 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 114.381 ms | 9 - 58 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 110.089 ms | 3 - 49 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 111.674 ms | 9 - 28 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 107.49 ms | 3 - 17 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 110.766 ms | 19 - 52 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 83.455 ms | 8 - 67 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 82.171 ms | 3 - 57 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 83.613 ms | 23 - 82 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 96.464 ms | 9 - 60 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 91.582 ms | 3 - 55 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 94.185 ms | 26 - 73 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 115.101 ms | 3 - 3 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 118.39 ms | 33 - 33 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 21.798 ms | 2 - 36 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 14.19 ms | 2 - 46 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 11.988 ms | 2 - 15 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 12.736 ms | 2 - 43 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 43.445 ms | 2 - 68 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 21.798 ms | 2 - 36 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 12.016 ms | 2 - 16 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 14.34 ms | 2 - 39 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 12.011 ms | 2 - 14 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 12.736 ms | 2 - 43 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 12.041 ms | 0 - 17 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 39.266 ms | 28 - 107 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 7.936 ms | 2 - 48 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 28.437 ms | 30 - 290 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 7.37 ms | 2 - 43 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 28.693 ms | 27 - 239 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 13.124 ms | 0 - 0 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 44.478 ms | 58 - 58 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 23.138 ms | 2 - 40 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 12.524 ms | 1 - 33 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 13.256 ms | 2 - 48 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 8.019 ms | 0 - 43 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 12.842 ms | 2 - 12 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 6.899 ms | 1 - 13 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 13.477 ms | 2 - 41 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 7.549 ms | 1 - 34 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | TFLITE | 135.049 ms | 15 - 55 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 26.04 ms | 1 - 47 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | TFLITE | 427.955 ms | 1 - 39 MB | CPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 23.138 ms | 2 - 40 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 12.524 ms | 1 - 33 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 12.87 ms | 0 - 12 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 6.934 ms | 1 - 11 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 15.14 ms | 2 - 42 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 8.733 ms | 1 - 35 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 12.953 ms | 2 - 18 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 6.911 ms | 1 - 12 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 13.477 ms | 2 - 41 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 7.549 ms | 1 - 34 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 12.919 ms | 2 - 13 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN_DLC | 6.923 ms | 1 - 11 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 53.658 ms | 19 - 43 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 8.893 ms | 2 - 49 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 4.626 ms | 1 - 46 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 40.405 ms | 22 - 87 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 9.077 ms | 2 - 44 MB | NPU | Segformer-Base.tflite |
| Segformer-Base | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN_DLC | 3.465 ms | 1 - 40 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 39.888 ms | 21 - 82 MB | NPU | Segformer-Base.onnx |
| Segformer-Base | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 7.776 ms | 0 - 0 MB | NPU | Segformer-Base.dlc |
| Segformer-Base | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 59.784 ms | 29 - 29 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) : 218.1
Estimated peak memory usage (MB): [7, 56]
Total # Ops : 544
Compute Unit(s) : npu (544 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 (
.tfliteexport): This tutorial provides a guide to deploy the .tflite model in an Android application.QNN (
.soexport ): This sample app provides instructions on how to use the.soshared 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
- SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers
- Source Model Implementation
Community
- Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
- For questions or feedback please reach out to us.
- Downloads last month
- 31
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support