抄録
To ensure the reliability of AI systems, it is crucial to consider the reliability of the sensors they use. In
this paper, we will focus on the reliability of LiDAR, a key 3D sensor used in autonomous driving and
robotics. We will address two key aspects: the importance of high-resolution LiDAR for autonomous
vehicles, and the potential for LiDAR to be compromised through “spoofing” attacks. First, we will explore
the importance of high-resolution LiDAR for autonomous vehicles, and how it allows cars to fully
utilize surrounding 3D information. We will also discuss recent developments in “next-generation”
LiDARs from a circuit system perspective. Next, we will turn to the issue of LiDAR security, specifically
the threat of spoofing attacks. We will introduce research on this topic and discuss methods for defending
against such attacks. By examining these two key aspects of LiDAR reliability, we aim to shed
light on the challenges and solutions for achieving trustworthy AI.
