Deep neural networks for 2D LIDAR images classification



Presented To

Department of Electrical/Electronic Engineering

This work aims to explore different deep neural network architectures for object classification in 2D LiDAR traffic images. The images are recorded using aLiDAR sensor attached to the mask of a car. There are two sources of these images: real-life recordings created while driving through streets of Brno and recordings from a simulator. An essential part of this work is also the annotation of LiDAR images, which is different for real data and simulated data.
1 Introduction Theworldhasseenagreatadvanceinthefieldofautonomousvehicles in the past 15 years. In 2003, the well known Tesla, Inc. company— formerly Tesla Motors—was founded in Delaware. In July 2004 The Stanford Racing Team [1] was created under the leadership of professor Sebastian Thrun, who is considered to be the father of self-driving cars. TheteammanagedtowintheDARPAGrandChallengein2005[2] using Stanley, an autonomousVolkswagenTouareg,whichcanbeseen at the Smithsonian’s National Museum of American History in Washington, DC. In 2009 the Google Self-Driving Car Project was founded, lead by Sebastian Thrun, and later became the Waymo company. Both, Tesla and Waymo, specialize in autonomous car developmentbutthey achieve it through different means. Tesla relies on many cameras, radar, and ultrasonics, while Waymo equips its vehicles with 3DVelodyne LiDAR (Light Detection and Ranging) sensor as well as cameras and radar. Both of these approaches require an understanding and fusion of the data from different sensors in real time, which requires a lot of computational power as well as efficient computational models. One of these models are neural networks which are used to segment images from cameras so that the vehicle can comprehend captured camera images and react to them. This is useful, for example, for tracking road surface markings, recognizing traffic lights and traffic signs or detecting other vehicles, pedestrians, etc. In this thesis, I explore the possibility of using neural networks for segmenting 2D LiDAR images for a software company ARTIN and its project Roboauto, which has provided me with recorded LiDAR images.Therearetwosourcesoftheseimages:real-worldenvironment and a simulator. As a part of this thesis, I also process and annotate datafromtherecordings.Icomparedifferentarchitecturesofnetworks trained either on real or simulated data. I also analyze the impact of substituting a part of these networks for a part of an autoencoder which was pretrained on bigger dataset. In Section 2 I present some information about the Roboauto project relevant to this thesis. In Section 3 I introduce the neural networks and also the most common technique for training them -the gradient descent algorithm with backpropagation. In Section 4 I explain the 1

About E-Project Material Centre

E-Project Material Centre is a web service aimed at successfully assisting final year students with quality, well researched, reliable and ready made project work. Our materials are recent, complete (chapter 1 to Minimum of Chapter 5, with references) and well written.INSTANT ACCESS! INSTANT DOWNLOAD. Simply select your department, choose from our list of topics available and explore your data

Why Students Love to Use E-Project Material ?

Guaranteed Delivery Getting your project delivered on time is essential. You cannot afford to turn in your project past the deadline. That is why you must get your project online from a company that guarantees to meet your deadline. e-Project Topics Material Centre is happy to offer instant delivery of projects listed on our website. We can handle just about any deadline you send our way. Satisfaction Guaranteed We always do whatever is necessary to ensure every customer's satisfaction


E-Project Topics Material Centre will only provide projects as a reference for your research. The projects ordered and produced should be used as a guide or framework for your own project. The contents of the projects should be able to help you in generating new ideas and thoughts for your own project. It is the aim of e-Project Topics Centre to only provide guidance by which the projects should be pursued. We are neither encouraging any form of plagiarism nor are we advocating the use of the projects produced herein for cheating.

Terms and Condition

Using our service is LEGAL and IS NOT prohibited by any university/college policies You are allowed to use the original model papers you will receive in the following ways:
  • As a source for additional understanding of the subject
  • As a source for ideas for you own research (if properly referenced)
  • For PROPER paraphrasing ( see your university definition of plagiarism and acceptable paraphrase) Direct citing ( if referenced properly)
Thank you so much for your respect to the authors copyright