Adversarial Style Augmentation for Domain Generalized Urban-Scene Segmentation

07/11/2022
by   Zhun Zhong, et al.
0

In this paper, we consider the problem of domain generalization in semantic segmentation, which aims to learn a robust model using only labeled synthetic (source) data. The model is expected to perform well on unseen real (target) domains. Our study finds that the image style variation can largely influence the model's performance and the style features can be well represented by the channel-wise mean and standard deviation of images. Inspired by this, we propose a novel adversarial style augmentation (AdvStyle) approach, which can dynamically generate hard stylized images during training and thus can effectively prevent the model from overfitting on the source domain. Specifically, AdvStyle regards the style feature as a learnable parameter and updates it by adversarial training. The learned adversarial style feature is used to construct an adversarial image for robust model training. AdvStyle is easy to implement and can be readily applied to different models. Experiments on two synthetic-to-real semantic segmentation benchmarks demonstrate that AdvStyle can significantly improve the model performance on unseen real domains and show that we can achieve the state of the art. Moreover, AdvStyle can be employed to domain generalized image classification and produces a clear improvement on the considered datasets.

READ FULL TEXT

page 2

page 3

page 5

page 9

page 17

page 18

page 22

page 23

research
04/06/2022

Style-Hallucinated Dual Consistency Learning for Domain Generalized Semantic Segmentation

In this paper, we study the task of synthetic-to-real domain generalized...
research
11/30/2017

ROAD: Reality Oriented Adaptation for Semantic Segmentation of Urban Scenes

Exploiting synthetic data to learn deep models has attracted increasing ...
research
05/30/2018

Generalizing to Unseen Domains via Adversarial Data Augmentation

We are concerned with learning models that generalize well to different ...
research
02/26/2022

DGSS : Domain Generalized Semantic Segmentation using Iterative Style Mining and Latent Representation Alignment

Semantic segmentation algorithms require access to well-annotated datase...
research
07/04/2023

Augment Features Beyond Color for Domain Generalized Segmentation

Domain generalized semantic segmentation (DGSS) is an essential but high...
research
04/22/2023

Improving Stain Invariance of CNNs for Segmentation by Fusing Channel Attention and Domain-Adversarial Training

Variability in staining protocols, such as different slide preparation t...
research
08/25/2023

A Re-Parameterized Vision Transformer (ReVT) for Domain-Generalized Semantic Segmentation

The task of semantic segmentation requires a model to assign semantic la...

Please sign up or login with your details

Forgot password? Click here to reset