文献类型：期刊文献

中文题名：基于多层级领域自适应网络和置信度约束的道路场景语义分割方法

英文题名：Semantic Segmentation Methods for Road Scenes Based on Multi-Level Domain Adaptation Network and Confidence Constraints

作者：万才路[1];堵威[1]

机构：[1]华东理工大学能源化工过程智能制造教育部重点实验室,上海200237

年份：2025

卷号：51

期号：3

起止页码：363

中文期刊名：华东理工大学学报(自然科学版)

外文期刊名：Journal of East China University of Science and Technology

收录：;北大核心:【北大核心2023】；

语种：中文

中文关键词：道路场景;语义分割;领域自适应;自训练;对抗学习

外文关键词：road scene;semantic segmentation;unsupervised domain adaptation;self-training;adversarial learning

摘要：语义分割旨在为图像中的每个像素分配一个类别标签,在自动驾驶领域具有广泛的应用。在实际应用中,在某个场景训练的语义分割模型无法有效地应用到其他场景,这成为实际应用中的一个关键问题,像素级域内自适应方法已被证明是一种解决此问题的有效方法。然而,这种方法不能有效地利用空间位置信息,并且容易受到噪声伪标签的影响。为了解决这些问题,本文首先提出了多层级领域自适应网络和置信度约束的方法,同时利用空间先验知识提出了阈值方法,提高了在域内自适应中使用伪标签的质量。在“GTA5到Cityscapes”和“SYNTHIA到Cityscapes”任务中,相较于基准方法本文所提方法分别实现了6.5%和2.8%的性能提升。
Semantic segmentation aims to assign a class label to each pixel in an image and has a wide range of applications.Semantic Segmentation needs large numbers of high-quality labels,which requires a lot of manpower and material resources.Furthermore,a semantic segmentation model trained on one domain cannot generalize well to other domains,which becomes a key problem in its practical applications.Unsupervised pixel-level intra-domain adaptation for semantic segmentation has been proven to be an effective method to address the problem.However,this method cannot effectively exploit spatial location information and is adversely affected by noisy pseudo-labels.In this work,we propose a confidence-guided multi-level domain adaptation approach to solve the problem.Specifically,we propose a multi-level domain adaptation framework to reduce the differences between pixels and spatial location information of images simultaneously.Moreover,to avoid that overfitting pseudo-labels may degrade the performance of the segmentation network,we construct a confidence loss function to constrain the network training.And we propose a method of selecting pseudo-labels and achieving better results in acquiring high-quality pseudo-labels than existing methods.We demonstrate the effectiveness of our approach through synthetic-to-real adaptation experiments.Compared with the unsupervised pixel-level intra-domain adaptation for semantic segmentation,our method leads to 6.5%and 2.8%relative improvements in mean intersection-over-union on the tasks“GTA5 to Cityscapes”and“SYNTHIA to Cityscapes”,respectively.