CitySemSegFormer

CitySemSegFormer Model Card

Model Overview

The model described in this card segments cityscapes urban city classes within an image and returns a semantic segmentation mask.

Model Architecture

Segformer is a real-time state of the art transformer based semantic segmentation model. SegFormer is a simple, efficient yet powerful semantic segmentation framework which unifies Transformers with lightweight multilayer perception (MLP) decoders. It then predicts a class label for every pixel in the input image. This model segments the urban cityscapes 19 classes which include:

road
sidewalk
building
wall
fence
pole
traffic light
traffic sign
vegetation
terrain
sky
person
rider
car
truck
bus
train
motorcycle
bicycle

Training Algorithm

The training algorithm optimizes the network to minimize the cross-entropy loss for every pixel of the mask.

Training Data

Citysemsegformer model was trained on a proprietary dataset with more than 2 million objects for car class. Most of the training dataset was collected in-house from images from a variety of dashcams and a small seed dataset containing images from traffic cameras in a city in the US. This content was chosen to auto-label urban city classes with segmentation masks. The approximate frequency distribution of predominant classes in the dataset are as following:

		Object	Distribution
Environment	Images	Cars	Persons	Road Signs	Two-Wheelers
Dashcam (5ft height)	128,000	1.7M	720,000	354,127	54,000
Traffic signal content	50,000	1.1M	53500	184000	11000
Total	178,000	2.8M	773,500	538,127	65,000

Training Data Ground-truth Labeling Guidelines

All objects were auto-labelled using NVSeg[2] model from NVIDIA as starting point.

Performance

Evaluation Data

The inference performance of CitySemSegFormer model was measured against 300 proprietary images that were hand labelled across a variety of environments. The frames are high resolution images 1920x1080 pixels resized to 1820x1024 pixels before passing to the CitySemSegFormer model.

Methodology and KPI

The KPI for the evaluation data are reported in the table below. Model is evaluated based on Mean Intersection-Over-Union. Mean Intersection-Over-Union (MIOU) is a common evaluation metric for semantic image segmentation, which first computes the IOU for each semantic class and then computes the average over classes.

Model	CitySemSegFormer
Classes	MIOU
road	65.34
sidewalk	13.92
building	72.45
wall	51.33
fence	65.57
pole	43.45
traffic light	8.56
traffic sign	72.8
vegetation	79.72
terrain	38.52
sky	96.47
person	81.32
rider	21.24
car	90.13
truck	33.12
bus	93.93
train	3.2
motorcycle	33.44
bicycle	39.94

Real-time Inference Performance

The inference is run on the provided unpruned models at INT8 precision. On the Jetson Nano FP16 precision is used. The inference performance is run using trtexec on Jetson Nano, AGX Xavier, Xavier NX and NVIDIA T4 GPU. The Jetson devices are running at Max-N configuration for maximum GPU frequency. The performance shown here is the inference only performance. The end-to-end performance with streaming video data might slightly vary depending on other bottlenecks in the hardware and software.

BS - Batch Size

		Xavier NX		AGX Xavier		Orin NX 16GB		Orin 64GB		T4		A100		A30		A10		A2
Model arch	Precision	GPU	2*DLA	GPU	2*DLA	GPU	2*DLA	GPU	2*DLA	BS	FPS	BS	FPS	BS	FPS	BS	FPS	BS	FPS
CitySemSegFormer	FP16	0.4	--	0.7	--	0.6	--	1.5	--	1	2	4	13	1	6.4	1	4.5	1	1.3

How to use this model

These models need to be used with NVIDIA Hardware and Software. For Hardware, the models can run on any NVIDIA GPU including NVIDIA Jetson devices. These models can only be used with Train Adapt Optimize (TAO) Toolkit, DeepStream SDK or TensorRT.

The model is intended for primarily deploying and doing inference using DeepStream.

Primary use case intended for the model is segmenting urban city classes in a color (RGB) image. The model can be used to segment urban city transport/ setting from photos and videos by using appropriate video or image decoding and pre-processing. Note this model performs semantic segmentation and not instance based segmentation.

The model is encrypted and will only operate with the following key:

Model load key: tlt_encode

Input

Color Images of resolution 1820x1024x3

Output

Category label (person or background) for every pixel in the input image. Outputs a semantic of urban city classes for the input image.

Output image

Note: Please note that Citysemsegformer currently can only be used as deployable model in Deepstream. In the current version, citySemSegformer does not provide support for fine-tuning with TAO-Toolkit.

Instructions to deploy these models with DeepStream

To create the entire end-to-end video analytics application, deploy these models with DeepStream SDK. DeepStream SDK is a streaming analytics toolkit to accelerate building AI-based video analytics applications. DeepStream supports direct integration of these models into the deepstream sample app.

To deploy these models with DeepStream 6.1, please follow the instructions below:

Download and install DeepStream SDK. The installation instructions for DeepStream are provided in DeepStream development guide. The config files for the purpose-built models are located in:

/opt/nvidia/deepstream is the default DeepStream installation directory. This path will be different if you are installing in a different directory.

You will need 1 config files and 1 label file. These files are provided in [NVIDIA-AI-IOT](@todo : update).

pgie_tlt_config_citysemsegformer.txt - File to configure inference settings for CitySemSegformer
labels.txt - Label file with 19 classes

Convert the .etlt file to engine if you want to input the model as TRT engine. Otherwise, you can input the etlt model directly to Deepstream. In order to manually convert to TRT engine, follow the example command below:

TAO-Converter Commands

FP16

./tao-converter -k tlt_encode -p input,1x3x1024x1820,1x3x1024x1820,1x3x1024x1820 -t fp16 -e ./bs1_fp16.engine ./citySemSegFormer.etlt

Key Parameters in pgie_tlt_config_citysemsegformer.txt

# You can either provide the etlt model and key or trt engine obtained by using tao-converter
tlt-model-key=tlt_encode
# tlt-encoded-model=../../path/to/.etlt file
model-engine-file=../../path/to/trt_engine
net-scale-factor=0.01735207357279195
offsets=123.675;116.28;103.53
# Since the model input channel is 3, using RGB color format.
model-color-format=0
labelfile-path=./labels.txt
infer-dims=3;1024;1820
batch-size=1
## 0=FP32, 1=INT8, 2=FP16 mode
network-mode=2
interval=0
gie-unique-id=1
cluster-mode=2
## 0=Detector, 1=Classifier, 2=Semantic Segmentation, 3=Instance Segmentation, 100=Other
network-type=100
output-tensor-meta=1
num-detected-classes=20
segmentation-output-order=1
roi-top-offset=0
roi-bottom-offset=0
detected-min-w=0
detected-min-h=0
detected-max-w=0
detected-max-h=0

Run ds-tao-segmentation:

ds-tao-segmentation -c configs/segformer_tao/pgie_tlt_config_citysemsegformer.txt -i $DS_SRC_PATH/samples/streams/sample_720p.h264

Documentation to deploy with DeepStream is provided in "Deploying to DeepStream" chapter of TAO User Guide.

Limitations

Under-represented classes

NVIDIA Citysemsegformer model was trained to detect classes that are predominantly found in road transport setting. It relatively performs poorly on under-represented classes in our internal Intelligent Transport System dataset. Some of these classes include: rider, truck, train, motorcycle.

Model versions:

CitySemSegformer:

deployable_citysemsegformer_v1.0 - citySemSegformer model deployable to deepstream.

References

Citations

Xie, Enze, et al. "SegFormer: Simple and efficient design for semantic segmentation with transformers." Advances in Neural Information Processing Systems 34 (2021): 12077-12090.
https://github.com/NVIDIA/semantic-segmentation

Using TAO Pre-trained Models

Get TAO Container
Get other Purpose-built models from NGC model registry:

Technical blogs

Read the 2 part blog on training and optimizing 2D body pose estimation model with TAO - Part 1 | Part 2
Learn how to train real-time License plate detection and recognition app with TAO and DeepStream.
Model accuracy is extremely important, learn how you can achieve state of the art accuracy for classification and object detection models using TAO
Learn how to train Instance segmentation model using MaskRCNN with TAO
Learn how to improve INT8 accuracy using Quantization aware training(QAT) with TAO
Read the technical tutorial on how PeopleNet model can be trained with custom data using Transfer Learning Toolkit
Learn how to train and deploy real-time intelligent video analytics apps and services using DeepStream SDK

License

License to use this model is covered by the Model EULA. By downloading the unpruned or pruned version of the model, you accept the terms and conditions of these licenses

Ethical Considerations

Training and evaluation dataset mostly consists of North American content. An ideal training and evaluation dataset would additionally include content from other geographies.

NVIDIA’s platforms and application frameworks enable developers to build a wide array of AI applications. Consider potential algorithmic bias when choosing or creating the models being deployed. Work with the model’s developer to ensure that it meets the requirements for the relevant industry and use case; that the necessary instruction and documentation are provided to understand error rates, confidence intervals, and results; and that the model is being used under the conditions and in the manner intended.

CitySemSegFormer

Description

Publisher

Latest Version

Modified

Size