基于自注意和对抗学习的道路场景水体检测方法

王臣毅; 王欢; 孟策

doi:10.12263/DZXB.20210743

您当前的位置：

首页 >

文章列表页 >

基于自注意和对抗学习的道路场景水体检测方法

学术论文 | 更新时间：2025-12-08

- 基于自注意和对抗学习的道路场景水体检测方法
- Water Puddle Detection Method for Road Scene Based on Self-Attention and Adversarial Learning
- 电子学报 2023年51卷第8期页码：2213-2225
- 作者机构：
  
  南京理工大学计算机科学与技术学院，江苏南京 210094
- 作者简介：
  
  [ "王臣毅男，1997年6月出生，重庆人.现为南京理工大学计算机科学与工程学院硕士研究生.主要研究方向为计算机视觉、深度学习. E-mail: wcyjerry@qq.com" ]
  [ "王欢（通讯作者）男，1982年7月出生，江苏句容人.南京理工大学计算机科学与工程学院学院副教授.主要研究方向为模式识别、图像处理、红外目标检测、目标跟踪、机器人视觉导航和人工智能. Email: wanghuanphd@njust.edu.cn" ]
  [ "孟策男，1997年11月出生，江苏盐城人.南京理工大学计算机科学与工程学院硕士研究生.主要研究方向为计算机视觉、深度学习." ]
- 基金信息：
  
  国家自然科学基金(61703209)
- DOI：10.12263/DZXB.20210743
  中图分类号： TP391.4;
- 收稿：2021-06-10，
  
  修回：2022-02-26，
  
  纸质出版：2023-08-25
- 稿件说明：
移动端阅览
王臣毅,王欢,孟策.基于自注意和对抗学习的道路场景水体检测方法[J].电子学报,2023,51(08):2213-2225.

WANG Chen-yi,WANG Huan,MENG Ce.Water Puddle Detection Method for Road Scene Based on Self-Attention and Adversarial Learning[J].ACTA ELECTRONICA SINICA,2023,51(08):2213-2225.
王臣毅,王欢,孟策.基于自注意和对抗学习的道路场景水体检测方法[J].电子学报,2023,51(08):2213-2225. DOI： 10.12263/DZXB.20210743.

WANG Chen-yi,WANG Huan,MENG Ce.Water Puddle Detection Method for Road Scene Based on Self-Attention and Adversarial Learning[J].ACTA ELECTRONICA SINICA,2023,51(08):2213-2225. DOI： 10.12263/DZXB.20210743.

摘要

无人驾驶车辆近年来一直是研究的热点.无人车运行环境复杂、不确定因素多，尤其当其意外驶入水坑、泥潭等地形时可能直接导致抛锚，造成不可估量的损失，因此水体检测对无人车的运行有着重要意义.借助深度网络的强大学习能力，本文首先将反射注意力单元和自注意力机制相结合，并在（U shape Network，U-Net）模型基础上添加残差卷积块和上采样卷积模块，得到了新的道路场景水体区域检测模型（U shape Network with Attention for Road，URA-Net），该模型能够更好地捕捉特征依赖关系，提高水体语义特征的表示能力.进一步，本文提出了一种基于双生成器对抗学习的训练模型（Redundant With Dual Generative Adversarial Network，RWD-GAN），它对URA-Net稍做修改，拓展成两个生成器，通过在对抗网络框架下让生成器与鉴别器、生成器与生成器之间实现对抗学习，促进不同网络模型之间的信息传递.在公开数据集上的大量实验表明URA-Net达到了87.18%的F1指标，而RWD-GAN模型能够进一步提高水体检测的精度，使提升到了88.54%，URA-Net和RWD-GAN均超出现有深度网络水体检测方法的性能表现.

Abstract

There has been much interest in the research of self-driving cars. Yet the detection of potentially dangerous obstacles on road makes this investigation more challenging. Water puddles

a typical obstacle of this kind

could trap a self-driving car or even cause serious accidents. Therefore

detecting water puddles is of great importance. To this end

this paper propose a novel water puddle detection model

URA-net(U shape Network with Attention for Road). Building its backbone on U-net(U shape Network) with residual and upsampling blocks added

URA-net combines both the reflection attention units and self-attention units

which can better characterize the dependence among image features so as to improve the representative capability to locate water puddles. Furthermore

a two-generator conditional adversarial network RWD-GAN(Redundant With Dual Generative Adversarial Network) is proposed

where two URA-Nets with a minor revision become the two generators to facilitate the information interaction in the adversarial learning process between the generators and the discriminator

as well as between the two generators themselves. Experiments on the public water puddle dataset demonstration that URA-net achieves 87.18% measure

while RWD-GAN can further improve the accuracy of URA-net

pushing F1-score to 88.54%. Both URA-net and RWD-GAN outperforms the state-of-the-arts.

关键词

Keywords

references

MATTHIES L H , BELLUTA P , MCHENRY M . Detecting water hazards for autonomous off-road navigation [C ] // Proceedings of SPIE , Unmanned Ground Vehicle Technology . Orlando : SPIE , 2003 , 231 - 242 .

ZHAO Y , DENG Y , PAN C , et al . Research of water hazard detection based on color and texture features [J ] . Sensors & Transducers , 2013 , 157 ( 10 ): 428 - 433 .

KIM J , BAEK J , CHOI H , et al . Wet area and puddle detection for advanced driver assistance systems (ADAS) using a stereo camera [J ] . International Journal of Control, Automation and Systems , 2016 , 14 ( 1 ): 263 - 271 .

YAO T , XIANG Z , LIU J , et al . Multi-feature fusion based outdoor water hazards detection [C ] // 2007 International Conference on Mechatronics and Automation . Harbin : IEEE , 2007 : 652 - 656 .

RANKIN A L , MATTHIES L H , BELLUTTA P . Daytime water detection based on sky reflections [C ] // 2011 IEEE International Conference on Robotics and Automation . Shanghai : IEEE , 2011 : 5329 - 5336 .

XIE B , PAN H , XIANG Z , et al . Polarization-based water hazards detection for autonomous off-road navigation [C ] // 2007 International Conference on Mechatronics and Automation . Harbin : IEEE , 2007 : 1666 - 1670 .

NGUYEN C V , MILFORD M , MAHONY R . 3D tracking of water hazards with polarized stereo cameras [C ] // 2017 IEEE International Conference on Robotics and Automation (ICRA) . Singapore : IEEE , 2017 : 5251 - 5257 .

SANTANA P , MENDONÇA R , BARATA J . Water detection with segmentation guided dynamic texture recognition [C ] // 2012 IEEE International Conference on Robotics and Biomimetics (ROBIO) . Guangzhou : IEEE , 2012 : 1836 - 1841 .

METTES P , TAN R T , VELTKAMP R . On the segmentation and classification of water in videos [C ] // 2014 . International Conference on Computer Vision Theory and Applications (VISAPP) . Lisbon : IEEE , 2014, 1 : 283 - 292 .

HAN X , NGUYEN C , YOU S , et al . Single image water hazard detection using fcn with reflection attention units [C ] // Proceedings of the European Conference on Computer Vision (ECCV) . Munich : IEEE , 2018 : 105 - 120 .

LONG J , SHELHAMER E , DARRELL T . Fully convolutional networks for semantic segmentation [C ] // Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition . Boston : IEEE , 2015 : 3431 - 3440 .

WANG L , WANG H . Water hzard detection using conditional generative adversarial network with mixture reflection attention units [J ] . IEEE Access , 2019 , 7 : 167497 - 167506 .

MIRZA M , OSINDERO S . Conditional generative adversarial nets [EB/OL ] . ( 2014-11-06 )[ 2021-06-10 ] . https://arxiv.org/abs/1411.1784 https://arxiv.org/abs/1411.1784 .

RONNEBERGER O , FISCHER P , BROX T . U-Net: Convolutional networks for biomedical image segmentation [C ] // International Conference on Medical Image Computing and Computer-Assisted Intervention . Cham : Springer . 2015 : 234 - 241 .

马勇 , 成谢锋 , 唐振民 , 等 . 野外场景分析与水体识别新方法的研究 [J ] . 电子学报 , 2013 , 43 ( 7 ): 1419 - 1424 .

MA Y , CHENG X F , TANG Z M , et al . New methods for outdoor scene analysis and water body identification [J ] . Acta Electronica Sinica , 2013 , 43 ( 7 ): 1419 - 1424 . (in Chinese)

MNIH V , HEESS N , GRAVES A . Recurrent models of visual attention [C ] // Advances in Neural Information Processing Systems . Montreal : MIT Press , 2014 : 2204 - 2212 .

乔思波 , 庞善臣 , 王敏 , 等 . 基于残差混合注意力机制的脑部CT图像分类卷积神经网络模型 [J ] . 电子学报 , 2021 , 49 ( 5 ): 984 - 991 .

QIAO S B , PANG S C , WANG M , et al . A convolutional neural network for brain CT image classification based on residual hybrid attention mechanism [J ] . Acta Electronica Sinica , 2021 , 49 ( 5 ): 984 - 991 . (in Chinese)

李志欣 , 孙亚茹 . 双路注意力引导图卷积网络的关系抽取 [J ] . 电子学报 , 2021 , 49 ( 2 ): 315 - 323 .

LI Z X , SUN Y R . Dual attention guided graph convolutional networks for relation extraction [J ] . Acta Electronica Sinica , 2021 , 49 ( 2 ): 315 - 323 . (in Chinese)

VASWANI A , SHAZEER N , PARMAR N , et al . Attention is all you need [C ] // Advances in Neural Information Processing Systems . Long Beach : MIT Press , 2017 : 5998 - 6008 .

GOODFELLOW I , POUGET-ABADIE J , MIRZA M , et al . Generative adversarial nets [C ] // Advances in neural information processing systems . Montreal : MIT Press , 2014 : 2672 - 2680 .

NGUYEN V , YAGO VICENTE T F , ZHAO M , et al . Shadow detection with conditional generative adversarial networks [C ] // Proceedings of the IEEE International Conference on Computer Vision . Honolulu : IEEE , 2017 : 4510 - 4518 .

CHEN M H , HONG Y X , DENG C . Domain adaptation for semantic segmentation with maximum squares loss . Proceedings of the IEEE International Conference on Computer Vision. Seoul : IEEE , 2019 : 1476 - 1480

冯建周 , 宋沙沙 , 王元卓 , 等 . 基于改进注意力机制的实体关系抽取方法 [J ] . 电子学报 , 2019 , 47 ( 8 ): 1692 - 1700 .

FENG J Z , SONG S S , WANG Y Z , et al . Entity relation extraction based on improved attention mechanism [J ] . Acta Electronica Sinica , 2019 , 47 ( 8 ): 1692 - 1700 . (in Chinese)

Kasi V , Kartikeya S , Peng J . Offseg: A semantic segmentation ramework for off-road driving [EB/OL ] . ( 2021-03-23 )[ 2021-06-10 ] . https://arxiv.org/abs/2103.12417 https://arxiv.org/abs/2103.12417 .

ZHENG W , YAN L , WANG F Y , et al . Progressive knowledge-embedded unified perceptual parsing for scene understanding [C ] // 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) . Nashville : IEEE , 2021 : 1633 - 1642 .

LIN T Y , DOLLAR P , GIRSHICK R , et al . Feature pyramid networks for object detection [C ] // IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . Honolulu : IEEE , 2017 : 1210 - 1214 .

WANG J , SUN K , CHENG T , et al . Deep high-resolution representation learning for visual recognition [EB/OL ] . ( 2019-08-20 )[ 2021-06-10 ] . https://arxiv.org/abs/1908.07919 https://arxiv.org/abs/1908.07919 .

YU C Q , WANG J B , PENG C , et al . BiSeNet: Bilateral segmentation network for real-time semantic segmentation [C ] // Proceedings of the European Conference on Computer Vision (ECCV) . Munich : Springer , 2018 , 325 - 341 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

面向时序异常检测的可变视距多向扫描方法

基于稀疏平滑自蒸馏的差分隐私深度学习方法

基于非一般类算子融合方法及硬件架构设计

跨模态渐进式知识迁移SAR目标检测技术