中文版 | English
题名

深度学习感知不确定性条件下的自动驾驶安全行为决策

其他题名
SAFE MANEUVER DECISION-MAKING FOR AUTONOMOUS DRIVING UNDER UNCERTAINTIES IN DEEP LEARNING-ENABLED PERCEPTION
姓名
姓名拼音
LIU Bowen
学号
12032504
学位类型
硕士
学位专业
0809 电子科学与技术
学科门类/专业学位类别
08 工学
导师
郝祁
导师单位
计算机科学与工程系
论文答辩日期
2023-05-13
论文提交日期
2023-07-01
学位授予单位
南方科技大学
学位授予地点
深圳
摘要
无人驾驶车辆采用计算机视觉和深度学习等技术,能够在无需驾驶员操作的 情况下自主行驶,其有望提高城市交通效率、减少交通事故、和改善出行体验。自 动驾驶技术的核心包括感知、决策和路径规划等方面,它们依赖于先进的传感器、 算法和控制系统。近些年来,自动驾驶相关方面的研究逐渐增加。与此同时自动驾 驶车辆同样面临着安全性、可靠性等多重挑战。由于复杂的行车环境和传感器误 差等原因,基于深度学习的感知模型,其检测结果存在不确定性。而自动驾驶系 统作为一个高度耦合,上下贯通的系统,感知模型的不确定性势必会影响到车辆 机动决策等下游模块。最近,存在使用深度学习,并利用概率形式来定量表示神 经网络不确定性的方法,试图缓解复杂环境、传感器误差对检测结果造成的影响, 下游决策模块在进行行为决策时,应兼顾考虑到感知模型输出的概率信息。然而 如何定量估计深度学习感知的不确定性,并在此机制上设计安全决策机制仍然是 一个亟待解决的问题。本文提出了一个统一的框架,该框架可以对感知模型的不 确定性进行量化表示,并通过兼顾考虑这些不确定性生成安全驾驶决策,本文的 主要贡献有三点:(1) 提出了支持不确定性量化的深度学习感知模型,完成对环境 中车辆和车道线的实时、实例级别的不确定性量化评估;(2) 设计了基于自适应似 然网络的安全决策方法,实现感知不确定性到决策过程的传递,建立了感知不确 定性与决策安全之间的关联,在此基础上实现安全决策;(3) 基于典型复杂场景对 本文提出的方法进行实验验证。实验结果表明,在复杂交通场景下,我们的机动 决策方法可以在效率和安全性方面取得优异的性能。
关键词
语种
中文
培养类别
独立培养
入学年份
2020
学位授予年份
2023-06
参考文献列表

[1] 全国机动车保有量达 4.17 亿辆驾驶人超过 5 亿人[J]. 道路交通管理, 2023(1): 1.
[2] 无. 我国首个国家级自动驾驶路测管理规范出台[J]. 汽车纵横, 2018(5): 2.
[3] 陆叶, 陈嘟, 何剑. 毫米波雷达在汽车自动驾驶中的应用与展望[Z]. 2020.
[4] 戴永江. 激光雷达原理[M]. 激光雷达原理, 2002.
[5] 曹建峰. 自动驾驶汽车” 撞人”, 谁来担责[J]. 方圆, 2018(7): 3.
[6] URICAR M, KRIZEK P, SISTU G, et al. SoilingNet: Soiling Detection on AutomotiveSurround-View Cameras[M]. IEEE, 2019.
[7] PORAV H, BRULS T, NEWMAN P. I Can See Clearly Now : Image Restoration via De-Raining[J]. IEEE, 2019.
[8] NAH S, KIM T H, LEE K M. Deep Multi-scale Convolutional Neural Network for DynamicScene Deblurring[J]. IEEE Computer Society, 2016.
[9] JO Y, OH S W, KANG J, et al. Deep Video Super-Resolution Network Using Dynamic Upsampling Filters Without Explicit Motion Compensation[C]//2018 IEEE/CVF Conference onComputer Vision and Pattern Recognition (CVPR). 2018.
[10] FENG D, HARAKEH A, WASLANDER S L, et al. A Review and Comparative Study onProbabilistic Object Detection in Autonomous Driving[J]. Institute of Electrical and ElectronicsEngineers (IEEE), 2021.
[11] GAWLIKOWSKI J, TASSI C, ALI M, et al. A Survey of Uncertainty in Deep Neural Networks.[Z]. 2021.
[12] YANG J, ZHOU K, LI Y, et al. Generalized Out-of-Distribution Detection: A Survey[Z]. 2021.
[13] KENDALL A, GAL Y. What uncertainties do we need in bayesian deep learning for computervision?[J]. Advances in neural information processing systems, 2017, 30.
[14] BAI H, CAI S, YE N, et al. Intention-aware online POMDP planning for autonomous driving ina crowd[C]//2015 ieee international conference on robotics and automation (icra). IEEE, 2015:454-460.
[15] CHEESEMAN P, GEVARTER W. Introduction to artificial intelligence[J]. Acm Sigart Bulletin,1986.
[16] 中国制造 2025[J]. 船舶标准化与质量, 2015(3): 2-15.
[17] VIOLA P, JONES M. Rapid object detection using a boosted cascade of simple features[C]//Proceedings of the 2001 IEEE computer society conference on computer vision and patternrecognition. CVPR 2001: volume 1. Ieee, 2001: I-I.
[18] FRIEDMAN J, HASTIE T, TIBSHIRANI R. Additive Logistic Regression: A Statistical Viewof Boosting[J]. The Annals of Statistics, 2000, 28(2): 337-407.
[19] DALAL N, TRIGGS B. Histograms of oriented gradients for human detection[C]//2005 IEEEcomputer society conference on computer vision and pattern recognition (CVPR’05): volume 1.Ieee, 2005: 886-893.
[20] GILBERT R C, TRAFALIS T B, ADRIANTO I. Support Vector Machines for Classification[M]. Encyclopedia of Operations Research and Management Science, 2011.
[21] UIJLINGS J R, VAN DE SANDE K E, GEVERS T, et al. Selective search for object recognition[J]. International journal of computer vision, 2013, 104: 154-171.
[22] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: Unified, real-time objectdetection[C]//Proceedings of the IEEE conference on computer vision and pattern recognition.2016: 779-788.
[23] LIU W, ANGUELOV D, ERHAN D, et al. Ssd: Single shot multibox detector[C]//ComputerVision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14,2016, Proceedings, Part I 14. Springer, 2016: 21-37.
[24] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedingsof the IEEE international conference on computer vision. 2017: 2980-2988.
[25] GIRSHICK R. Fast r-cnn[C]//Proceedings of the IEEE international conference on computervision. 2015: 1440-1448.
[26] REN S, HE K, GIRSHICK R, et al. Faster r-cnn: Towards real-time object detection with regionproposal networks[J]. Advances in neural information processing systems, 2015, 28.
[27] TIAN Z, SHEN C, CHEN H, et al. Fcos: Fully convolutional one-stage object detection[C]//Proceedings of the IEEE/CVF international conference on computer vision. 2019: 9627-9636.
[28] WANG T, ZHU X, PANG J, et al. Fcos3d: Fully convolutional one-stage monocular 3d object detection[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2021: 913-922.
[29] DUAN K, XIE L, QI H, et al. Corner proposal network for anchor-free, two-stage object detection[C]//Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August23–28, 2020, Proceedings, Part III. Springer, 2020: 399-416.
[30] CHEN X, MA H, WAN J, et al. Multi-view 3d object detection network for autonomous driving[C]//Proceedings of the IEEE conference on Computer Vision and Pattern Recognition. 2017:1907-1915.
[31] YANG B, LUO W, URTASUN R. Pixor: Real-time 3d object detection from point clouds[C]//Proceedings of the IEEE conference on Computer Vision and Pattern Recognition. 2018:7652-7660.
[32] LANG A H, VORA S, CAESAR H, et al. Pointpillars: Fast encoders for object detectionfrom point clouds[C]//Proceedings of the IEEE/CVF conference on computer vision and patternrecognition. 2019: 12697-12705.
[33] YAN Y, MAO Y, LI B. Second: Sparsely embedded convolutional detection[J]. Sensors, 2018,18(10): 3337.
[34] QI C R, SU H, MO K, et al. Pointnet: Deep learning on point sets for 3d classification and segmentation[C]//Proceedings of the IEEE conference on computer vision and pattern recognition.2017: 652-660.
[35] QI C R, YI L, SU H, et al. Pointnet++: Deep hierarchical feature learning on point sets in ametric space[J]. Advances in neural information processing systems, 2017, 30.
[36] SHI S, WANG X, LI H. Pointrcnn: 3d object proposal generation and detection from pointcloud[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition.2019: 770-779.
[37] READING C, HARAKEH A, CHAE J, et al. Categorical depth distribution network for monocular 3d object detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision andPattern Recognition. 2021: 8555-8564.
[38] CANNY J. A computational approach to edge detection[J]. IEEE Transactions on patternanalysis and machine intelligence, 1986(6): 679-698.
[39] HOUGH P V. Method and means for recognizing complex patterns[M]. Google Patents, 1962.
[40] TUSIMPLE. Tusimple Benchmark[Z]. 2017.
[41] PAN X, SHI J, LUO P, et al. Spatial as deep: Spatial cnn for traffic scene understanding[C]//Proceedings of the AAAI Conference on Artificial Intelligence: volume 32. 2018.
[42] BEHRENDT K, SOUSSAN R. Unsupervised labeled lane markers using maps[C]//Proceedingsof the IEEE/CVF International Conference on Computer Vision Workshops. 2019: 0-0.
[43] TABELINI L, BERRIEL R, PAIXAO T M, et al. Keep your eyes on the lane: Real-timeattention-guided lane detection[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2021: 294-302.
[44] GUO Y, CHEN G, ZHAO P, et al. Gen-lanenet: A generalized and scalable approach for 3dlane detection[C]//Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK,August 23–28, 2020, Proceedings, Part XXI 16. Springer, 2020: 666-681.
[45] DENKER J, LECUN Y. Transforming neural-net output levels to probability distributions[J].Advances in neural information processing systems, 1990, 3.
[46] MACKAY D J. A practical Bayesian framework for backpropagation networks[J]. Neural computation, 1992, 4(3): 448-472.
[47] GRAVES A. Practical variational inference for neural networks[J]. Advances in neural information processing systems, 2011, 24.
[48] GAL Y, GHAHRAMANI Z. Dropout as a bayesian approximation: Representing model uncertainty in deep learning[C]//international conference on machine learning. PMLR, 2016: 1050-1059.
[49] SRIVASTAVA N, HINTON G, KRIZHEVSKY A, et al. Dropout: A Simple Way to PreventNeural Networks from Overfitting[J]. Journal of Machine Learning Research, 2014, 15(1):1929-1958.
[50] DING L, LI D, LIU B, et al. Capture Uncertainties in Deep Neural Networks for Safe Operationof Autonomous Driving Vehicles[C]//2021 IEEE Intl Conf on Parallel & Distributed Processingwith Applications, Big Data & Cloud Computing, Sustainable Computing & Communications,Social Computing & Networking (ISPA/BDCloud/SocialCom/SustainCom). IEEE, 2021: 826-835.
[51] PAN H, WANG Z, ZHAN W, et al. Towards better performance and more explainable uncertainty for 3d object detection of autonomous vehicles[C]//2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC). IEEE, 2020: 1-7.
[52] HUANG P Y, HSU W T, CHIU C Y, et al. Efficient uncertainty estimation for semantic segmentation in videos[C]//Proceedings of the European Conference on Computer Vision (ECCV).2018: 520-535.
[53] SENSOY M, KAPLAN L, KANDEMIR M. Evidential deep learning to quantify classificationuncertainty[J]. Advances in neural information processing systems, 2018, 31.
[54] LUO Y, MEGHJANI M, HO Q H, et al. Interactive planning for autonomous urban drivingin adversarial scenarios[C]//2021 IEEE International Conference on Robotics and Automation(ICRA). IEEE, 2021: 5261-5267.
[55] SCHUBERT R, SCHULZE K, WANIELIK G. Situation assessment for automatic lane-changemaneuvers[J]. IEEE Transactions on Intelligent Transportation Systems, 2010, 11(3): 607-616.
[56] SCHUBERT R, WANIELIK G. Empirical evaluation of a unified bayesian object and situationassessment approach for lane change assistance[C]//2011 14th International IEEE Conferenceon Intelligent Transportation Systems (ITSC). IEEE, 2011: 1471-1476.
[57] CAESAR H, BANKITI V, LANG A H, et al. nuScenes: A multimodal dataset for autonomousdriving[Z]. 2019.
[58] DOSOVITSKIY A, ROS G, CODEVILLA F, et al. CARLA: An Open Urban Driving Simulator[A]. 2017. arXiv: 1711.03938.
[59] IZMAILOV P, MADDOX W J, KIRICHENKO P, et al. Subspace Inference for BayesianDeep Learning[C/OL]//GLOBERSON A, SILVA R. Proceedings of Machine Learning Research: volume 115 Proceedings of the Thirty-Fifth Conference on Uncertainty in ArtificialIntelligence, UAI 2019, Tel Aviv, Israel, July 22-25, 2019. AUAI Press, 2019: 1169-1179.http://proceedings.mlr.press/v115/izmailov20a.html.
[60] NEAL R M. Bayesian Learning for Neural Networks[J]. ieee transactions on neural networks,1994.
[61] FOONG A Y K, LI Y, HERNÁNDEZ-LOBATO J M, et al. ’In-Between’ Uncertainty inBayesian Neural Networks[J/OL]. CoRR, 2019, abs/1906.11537. http://arxiv.org/abs/1906.11537.
[62] HÜLLERMEIER E, WAEGEMAN W. Aleatoric and epistemic uncertainty in machine learning: An introduction to concepts and methods[J]. Machine Learning, 2021, 110: 457-506.
[63] BELUCH W H, GENEWEIN T, NÜRNBERGER A, et al. The Power of Ensembles forActive Learning in Image Classification[C/OL]//2018 IEEE Conference on Computer Visionand Pattern Recognition, CVPR 2018, Salt Lake City, UT, USA, June 18-22, 2018. Computer Vision Foundation / IEEE Computer Society, 2018: 9368-9377. http://openaccess.thecvf.com/content_cvpr_2018/html/Beluch_The_Power_of_CVPR_2018_paper.html. DOI:10.1109/CVPR.2018.00976.
[64] GUSTAFSSON F K, DANELLJAN M, SCHÖN T B. Evaluating Scalable Bayesian DeepLearning Methods for Robust Computer Vision[C/OL]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR Workshops 2020, Seattle, WA, USA, June 14-19,2020. Computer Vision Foundation / IEEE, 2020: 1289-1298. https://openaccess.thecvf.com/content_CVPRW_2020/html/w20/Gustafsson_Evaluating_Scalable_Bayesian_Deep_Learning_Methods_for_Robust_Computer_Vision_CVPRW_2020_paper.html. DOI: 10.1109/CVPRW50498.2020.00167.
[65] BLUM H, SARLIN P, NIETO J I, et al. Fishyscapes: A Benchmark for Safe Semantic Segmentation in Autonomous Driving[C/OL]//2019 IEEE/CVF International Conference on ComputerVision Workshops, ICCV Workshops 2019, Seoul, Korea (South), October 27-28, 2019. IEEE,2019: 2403-2412. https://doi.org/10.1109/ICCVW.2019.00294.
[66] LAKSHMINARAYANAN B, PRITZEL A, BLUNDELL C. Simple and scalable predictiveuncertainty estimation using deep ensembles[J]. Advances in neural information processingsystems, 2017, 30.
[67] SNOEK J, OVADIA Y, FERTIG E, et al. Can you trust your model’s uncertainty? Evaluating predictive uncertainty under dataset shift[C/OL]//WALLACH H M, LAROCHELLE H,BEYGELZIMER A, et al. Advances in Neural Information Processing Systems 32: AnnualConference on Neural Information Processing Systems 2019, NeurIPS 2019, December 8-14,2019, Vancouver, BC, Canada. 2019: 13969-13980. https://proceedings.neurips.cc/paper/2019/hash/8558cb408c1d76621371888657d2eb1d-Abstract.html.
[68] GAST J, ROTH S. Lightweight Probabilistic Deep Networks[C/OL]//2018 IEEE Conference onComputer Vision and Pattern Recognition, CVPR 2018, Salt Lake City, UT, USA, June 18-22,2018. Computer Vision Foundation / IEEE Computer Society, 2018: 3369-3378. http://openaccess.thecvf.com/content_cvpr_2018/html/Gast_Lightweight_Probabilistic_Deep_CVPR_2018_paper.html. DOI: 10.1109/CVPR.2018.00355.
[69] PAN H, WANG Z, ZHAN W, et al. Towards Better Performance and More Explainable Uncertainty for 3D Object Detection of Autonomous Vehicles[M]. International Conference onIntelligent Transportation Systems, 2020.
[70] MEYER G P, LADDHA A, KEE E, et al. LaserNet: An Efficient Probabilistic 3D Object Detector for Autonomous Driving[C/OL]//IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2019, Long Beach, CA, USA, June 16-20, 2019. Computer Vision Foundation /IEEE, 2019: 12677-12686. http://openaccess.thecvf.com/content_CVPR_2019/html/Meyer_LaserNet_An_Efficient_Probabilistic_3D_Object_Detector_for_Autonomous_Driving_CVPR_2019_paper.html. DOI: 10.1109/CVPR.2019.01296.
[71] CHOI S, LEE K, LIM S, et al. Uncertainty-Aware Learning from Demonstration Using Mixture Density Networks with Sampling-Free Variance Modeling[C/OL]//2018 IEEE InternationalConference on Robotics and Automation, ICRA 2018, Brisbane, Australia, May 21-25, 2018.IEEE, 2018: 6915-6922. https://doi.org/10.1109/ICRA.2018.8462978.
[72] HE Y, ZHU C, WANG J, et al. Bounding box regression with uncertainty for accurate object detection[C]//Proceedings of the ieee/cvf conference on computer vision and pattern recognition.2019: 2888-2897.
[73] YUJUN ZHANG W F H F M W Q L C L, Lei Zhu, WANG S. VIL-100: A New Dataset and ABaseline Model for Video Instance Lane Detection[C]//The IEEE International Conference onComputer Vision (ICCV). 2021.
[74] BAHDANAU D, CHO K, BENGIO Y. Neural Machine Translation by Jointly Learning to Alignand Translate[J]. Computer Science, 2014.
[75] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//Proceedingsof the IEEE conference on computer vision and pattern recognition. 2016: 770-778.
[76] GLOROT X, BORDES A, BENGIO Y. Deep sparse rectifier neural networks[C]//Proceedingsof the fourteenth international conference on artificial intelligence and statistics. JMLR Workshop and Conference Proceedings, 2011: 315-323.
[77] HUPFER C. Deceleration to safety time (DST)-a useful figure to evaluate traffic safety[C]//ICTCT Conference Proceedings of Seminar: volume 3. 1997: 5-7.
[78] SHAH S, DEY D, LOVETT C, et al. AirSim: High-Fidelity Visual and Physical Simulation forAutonomous Vehicles[A]. 2017. arXiv: 1705.05065.
[79] 张茜, 吴帆, 范文浩. 车联网仿真器 SUMO 介绍[Z]. 2020.

所在学位评定分委会
电子科学与技术
国内图书分类号
TP399
来源库
人工提交
成果类型学位论文
条目标识符http://sustech.caswiz.com/handle/2SGJ60CL/544761
专题工学院_计算机科学与工程系
推荐引用方式
GB/T 7714
刘博文. 深度学习感知不确定性条件下的自动驾驶安全行为决策[D]. 深圳. 南方科技大学,2023.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可 操作
12032504-刘博文-计算机科学与工(11822KB)----限制开放--请求全文
个性服务
原文链接
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
导出为Excel格式
导出为Csv格式
Altmetrics Score
谷歌学术
谷歌学术中相似的文章
[刘博文]的文章
百度学术
百度学术中相似的文章
[刘博文]的文章
必应学术
必应学术中相似的文章
[刘博文]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
[发表评论/异议/意见]
暂无评论

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。