carmaker软件_值得收藏！无人驾驶软件领域各大模块课程资源最全整理（从入门到放弃）...

文章来自知乎 作者洪泽鑫

原文链接：https://zhuanlan.zhihu.com/p/134673230

无人驾驶，从软件的功能模块来看有这么几个，给打算自学且正在选择方向的一个参考：

ー 1 ー

地图&定位

SLAM是自动驾驶最先在做的领域了。在建图这一块，前几年还在给大家普及什么叫高精度地图，但现在传统的图商也能又快又好地做出来一张高精度地图了，就是标准不一致比较难搞。

要学SLAM，有几个必备的资料：

• Cyrill Stachniss的youtube线上课程和课件

  https://www.youtube.com/watch?t=246s&v=wVsfCnyt5jA

  http://ais.informatik.uni-freiburg.de/teaching/ws13/mapping/

• Sebastian Thrun的书Probabilistic Robotics

  https://www.amazon.com/Probabilistic-Robotics-INTELLIGENT-ROBOTICS-AUTONOMOUS/dp/0262201623

• SLAM: The Essential Algorithms

  http://www-personal.acfr.usyd.edu.au/tbailey/publications/slamtutorial1.htm

• A practical introduction to pose-graph SLAM with ROS

  https://www.sauravag.com/2017/07/an-practical-introduction-to-pose-graph-slam/

• Probabilistic Systems Analysis and Applied Probability

  https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-041-probabilistic-systems-analysis-and-applied-probability-fall-2010/

以及，多看点最新的paper，试下最新的算法。 ー 2 ー

感知

感知是最基础的，也是业内前几年在重点发力的模块，目前感知单靠计算机视觉或者单靠雷达都是不可靠的，所以要学就学最核心的一块——多传感器融合和时序融合。 前几年搞算法是最吃香的，算法研发的薪资简直高到离谱，但这个领域的门槛是很高的，如果不是手握多篇顶会paper，就不要轻易尝试了，前几年也一下子被资本催生得过火了。 多传感器融合和时序融合先要会C++，列了些入门的免费课程间 ：

• C++ Tutorial for Complete Beginners(Udemy)
• Learn C++ by Creating(Udemy)
• C++, Short, and Sweet, Part 1(Udemy)
• C++ Programming Basics(Udemy)
• Beginning C++ Templates(Udemy)

C++学得差不多，就该打ACM了

• How to prepare for ACM – ICPC?https://www.geeksforgeeks.org/how-to-prepare-for-acm-icpc/

对实时定位也要懂(不懂的翻回去看上面) 了解下系统动力学(System dynamics)

• Introduction to System Dynamicshttps://ocw.mit.edu/courses/sloan-school-of-management/15-871-introduction-to-system-dynamics-fall-2013/

以及运动学建模(kinematics modeling)

• Kinematic Model - an overviewhttps://www.sciencedirect.com/topics/engineering/kinematic-model
• Kinematics: Describing the Motions of Spacecraft (Coursera)

多传感器融合相关资料：

• Handbook of Multisensor Data Fusion: Theory and Practice, Second Editionhttps://books.google.com.hk/books?id=7ZqiHyc-RhUC&pg=PR13&dq=Statistical+Sensor+Fusion.+Exercises.+pdf&hl=zh-CN&source=gbs_toc_r&cad=3#v=onepage&q&f=false
• Statistical Sensor Fusion. Fredrik Gustafsson. Studentlitteratur, 2018.https://www.amazon.com/Statistical-Sensor-Fusion-Fredrik-Gustafsson/dp/9144127243
• Statistical Sensor Fusion. Exercises. Christian Lundquist, Zoran Sjanic, and Fredrik Gustafsson. Studentlitteratur, 2015.https://www.amazon.com/exec/obidos/ASIN/9144100116/acmorg-20
• Statistical Sensor Fusion Matlab Toolboxhttp://www.control.isy.liu.se/student/tsrt14/file/sfmanual.pdf
• Android app Sensor Fusion apphttps://play.google.com/store/apps/details?id=com.hiq.sensor

如果你还是对算法情有独钟，那就先从计算机视觉开始吧：

• Tombone's Computer Vision Blog: From feature descriptors to deep learning 20 years of computer visionhttps://www.computervisionblog.com/2015/01/from-feature-descriptors-to-deep.html
• Andrew Ng的Machine Learning课程 (Coursera)
• Convolutional Neural Networks for Visual Recognition
• http://cs231n.stanford.edu/
• Tombone's Computer Vision Bloghttps://www.computervisionblog.com/
• Andrej Karpathy bloghttp://karpathy.github.io/
• 实践下Google Vision APIhttps://cloud.google.com/vision/

ー 3 ー

仿真

为什么不把更重要的决策规划放在前面？因为仿真也是一个基础模块，自动驾驶仿真软件很多，比如Perscan、Carmaker、VTD，以及开源的Carla、Airsim等。 虽然现在业内更多公司用的基于游戏引擎做的仿真系统，但是看看Waymo的Carcraft，还是自己研发的仿真系统，个人感觉这个才是大势所趋。 必备技能C++ 全栈开发能力也需要很强，以下也是些免费的资料

• Learn X in Y Minuteshttps://learnxinyminutes.com/
• What CS Majors Should Knowhttp://matt.might.net/articles/what-cs-majors-should-know/
• Google’s Technical Development Guidehttps://techdevguide.withgoogle.com/
• CSS Tricks — Complete Flexbox Guidehttps://css-tricks.com/snippets/css/a-guide-to-flexbox/
• Regex Cheat Sheethttps://regexlib.com/(X(1)A(fIuA8Yg0SbS8CAeDi6nMLdWQVg-n9Q-6jA2mfTBM0-rid3KaXbLK9xGATn2nyPuPdkiiSGxK0EHcn2Wp4Oh1uOsy7VIHbdk1RzfoNozSK0PnrN3lxy40EERySo5uSJOiioVFV_WogkAhYx4zywn5pcwlHiXSLKv9uPgI8-kM-Qr-rQO-u76gj7pIWTaGzumD0))/CheatSheet.aspx?AspxAutoDetectCookieSupport=1
• DevDocshttps://devdocs.io/
• Awesome List of Everything Programminghttps://github.com/sindresorhus/awesome
• How to Break Into the Tech Industry — a Guide to Job Hunting and Tech Interviewshttps://haseebq.com/how-to-break-into-tech-job-hunting-and-interviews/

学一下AngularJS 或 Angular2

• AngularJS For Beginnershttps://www.udemy.com/course/angularjs-for-beginners-udemy/
• AngularJS: Framework Fundamentalshttps://www.classcentral.com/course/edx-angularjs-framework-fundamentals-7377

机器学习也还得略懂一二

• Free Machine Learning Course (fast.ai)
• Machine Learning Course by Stanford University (Coursera)
• Deep Learning Course (deeplearning.ai)
• Machine Learning Course A-Z™: Hands-On Python & R In Data Science (Udemy)
• Free Machine Learning Data Science Course (Harvard University)

ー 4 ー

运动决策规划

这是目前在重点攻克的难题，可以说也是潜力最大的领域。许多人都是从机器人转过来的。 重要的事说三遍：C++，C++，C++ ROS入门

• Tutorials : Installationhttp://gazebosim.org/tutorials?cat=install
• Powering the world's robotshttps://www.ros.org/
• UBC Open Roboticshttps://github.com/openrobotics
• Learn TurtleBot and ROShttp://learn.turtlebot.com/
• Husky UR5 Mobile Manipulation Demohttp://wiki.ros.org/husky_ur5_moveit_config/Tutorials/Husky%20UR5%20Mobile%20Manipulation%20Demo

重点了解下运动优化(Trajectory optimization)

• Introduction to Trajectory Optimizationhttps://www.youtube.com/watch?v=wlkRYMVUZTs
• Overview of Trajectory Optimization Techniqueshttps://link.springer.com/chapter/10.1007/978-981-13-9845-2_2
• Underactuated Robotics - Lecture 9: Trajectory Optimizationhttps://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-832-underactuated-robotics-spring-2009/video-lectures/lecture-9-trajectory-optimization/
• An Overview of Planning Under Uncertainthttps://www.cs.nmsu.edu/~tson/classes/spring04-579/blythe-aimag.pdf

深度学习 也是一定要会的

• Neural Networks and Deep Learning-Andrew Ng(Coursera)
• How to Learn Machine Learning, The Self-Starter Wayhttps://elitedatascience.com/learn-machine-learning
• Jeremy Howard: The wonderful and terrifying implications of computers that can learnhttps://www.ted.com/talks/jeremy_howard_the_wonderful_and_terrifying_implications_of_computers_that_can_learn
• Blaise Agüera y Arcas: How computers are learning to be creativehttps://www.ted.com/talks/blaise_aguera_y_arcas_how_computers_are_learning_to_be_creative
• Anthony Goldbloom: The jobs we'll lose to machines — and the ones we won'thttps://www.ted.com/talks/anthony_goldbloom_the_jobs_we_ll_lose_to_machines_and_the_ones_we_won_t

以上，还是先从挑一个方向开始吧。

---

也许你会想阅读....

一场自动驾驶事故的背后  ：责任与伦理白话自动驾驶 | 自动驾驶车的数据之谜总结自动驾驶路测的 12 大验证要点智能网联汽车的概念与网联化应用浅谈自动驾驶中的动态高清地图合集 | 自动驾驶的Keyword 合集 | 自动驾驶初创公司

---

— THE END —

   转载目的在于学习，如有侵权联系删除。