Figure 1: Overview of the data processing pipeline for LiDAR-based perception in Baidu Apollo.
Step 1: Pre processing
. The raw LiDAR sensor input is called
3D point cloud and we denote it as
X
. The dimension of
X
is
n ×
4,
where
n
denotes the number of data points and each data point
is a 4-dimension vector with the 3D coordinates,
w
x
,
w
y
, and
w
z
,
and the intensity of the point. In the pre-processing step,
X
is rst
transformed into an absolute coordinate system. Next, the Region
of Interest (ROI) lter removes unrelated portions of the 3D point
cloud data, e.g., those that are outside of the road, based on HDMap
information. Next, a feature generation process generates a feature
matrix
x (
8
×
512
×
512
)
, which is the input to the subsequent
machine learning model. In this process, the ROI-ltered 3D point
cloud within the range (60 meters by default) is mapped to 512
×
512
cells according to the
w
x
and
w
y
coordinates. In each cell, the
assigned points are used to generate 8 features as listed in Table 1.
Step 2: DNN-based object detection.
A Deep Neural Network
(DNN) then takes the feature matrix
x
as input and produces a set
of output metrics for each cell, e.g., the probability of the cell being
a part of an obstacle. These output metrics are listed in Table 2.
Step 3: Post processing.
The clustering process only considers
cells with objectness values (one of the output metrics listed in Table
2) greater than a given threshold (0
.
5 by default). Then, the process
constructs candidate object clusters by building a connected graph
using the cells’ output metrics. Candidate object clusters are then l-
tered by selecting clusters with average positiveness values (another
output metric) greater than a given threshold (0
.
1 by default). The
box builder then reconstructs the bounding box including height,
width, length of an obstacle candidate from the 3D point cloud
assigned to it. Finally, the tracker integrates consecutive frames of
processed results to generate tracked obstacles, augmented with
additional information such as speed, acceleration, and turning
rates, as the output of the LiDAR-based perception.
With the information of perceived obstacles such as their po-
sitions, shapes, and obstacle types, the Apollo system then uses
such information to make driving decisions. The perception output
is further processed by the prediction module which predicts the
future trajectories of perceived obstacles, and then the planning
module which plans the future driving routes and makes decisions
such as stopping, lane changing, yielding, etc.
Feature Description
Max height Maximum height of points in the cell.
Max intensity Intensity of the brightest point in the cell.
Mean height Mean height of points in the cell.
Mean intensity Mean intensity of points in the cell.
Count Number of points in the cell.
Direction Angle of the cell’s center with respect to the origin.
Distance Distance between the cell’s center and the origin.
Non-empty
Binary value indicating whether the cell is empty or occupied.
Table 1: DNN model input features.
Metrics Description
Center oset
Oset to the predicted center of the cluster the cell belongs
to.
Objectness The probability of a cell belonging to an obstacle.
Positiveness The condence score of the detection.
Object height The predicted object height.
Class probability
The probability of the cell being a part of a vehicle, pedes-
trian, etc.
Table 2: DNN model output metrics.
2.2 LiDAR Sensor and Spoong Attacks
To understand the principles underlying our security analysis
methodology, it is necessary to understand how the LiDAR sen-
sor generates a point cloud and how it is possible to alter it in a
controlled way using spoong attacks.
LiDAR sensor.
A LiDAR sensor functions by ring laser pulses
and capturing their reections using photodiodes. Because the
speed of light is constant, the time it takes for the echo pulses to
reach the receiving photodiode provides an accurate measurement
of the distance between a LiDAR and a potential obstacle. By ring
the laser pulses at many vertical and horizontal angles, a LiDAR
generates a point cloud used by the AV systems to detect objects.
LiDAR spoong attack.
Sensor spoong attacks use the same
physical channels as the targeted sensor to manipulate the sen-
sor readings. This strategy makes it very dicult for the sensor
system to recognize such attack, since the attack doesn’t require
any physical contact or tampering with the sensor, and it doesn’t
interfere with the processing and transmission of the digital sensor
measurement. These types of attack could trick the victim sensor
to provide seemingly legitimate but actually erroneous data.
LiDAR has been shown to be vulnerable to laser spoong attacks
in prior work. Petit et al. demonstrated that a LiDAR spoong
attack can be performed by replaying the LiDAR laser pulses from
a dierent position to create fake points further than the location of
the spoofer [
42
]. Shin et al. showed that it is possible to generate a
fake point cloud at dierent distances, even closer than the spoofer
location [
44
]. In this paper, we build upon these prior work to study
the eect of this attack vector on the security of AV perception.
2.3 Adversarial Machine Learning
Neural networks.
A neural network is a function consisting of
connected units called (articial) neurons that work together to rep-
resent a dierentiable function that outputs a distribution. A given
neural network (e.g., classication) can be dened by its model
architecture and parameters
ϕ
. An optimizer such as Adam [
35
]
is used to update the parameters
ϕ
with respect to the objective
function L.
Adversarial examples.
Given a machine learning model
M
, in-
put
x
and its corresponding label
y
, an adversarial attacker aims
Session 10A: Cyberphysical Security
CCS ’19, November 11–15, 2019, London, United Kingdom
2269
剩余14页未读，继续阅读
Shmily~_~
- 粉丝: 4
- 资源: 1
我的内容管理
展开
最新资源
- AirKiss技术详解：无线传递信息与智能家居连接
- Hibernate主键生成策略详解
- 操作系统实验：位示图法管理磁盘空闲空间
- JSON详解：数据交换的主流格式
- Win7安装Ubuntu双系统详细指南
- FPGA内部结构与工作原理探索
- 信用评分模型解析：WOE、IV与ROC
- 使用LVS+Keepalived构建高可用负载均衡集群
- 微信小程序驱动餐饮与服装业创新转型：便捷管理与低成本优势
- 机器学习入门指南：从基础到进阶
- 解决Win7 IIS配置错误500.22与0x80070032
- SQL-DFS：优化HDFS小文件存储的解决方案
- Hadoop、Hbase、Spark环境部署与主机配置详解
- Kisso：加密会话Cookie实现的单点登录SSO
- OpenCV读取与拼接多幅图像教程
- QT实战：轻松生成与解析JSON数据
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
