A perception centred self-driving system without HD Maps

TL;DR

This paper introduces a perception-centered self-driving system that operates without HD maps, GPS, or IMU, using a novel line detection method to localize and understand traffic features, enhancing scalability and adaptability.

Contribution

The paper presents a new localization approach based solely on traffic feature detection, eliminating the need for labor-intensive HD maps and improving scalability in autonomous driving.

Findings

Effective detection of diverse traffic lines demonstrated.

Localization accuracy achieved without HD maps or GPS.

Robustness tested across multiple datasets.

Abstract

Building a fully autonomous self-driving system has been discussed for more than 20 years yet remains unsolved. Previous systems have limited ability to scale. Their localization subsystem needs labor-intensive map recording for running in a new area, and the accuracy decreases after the changes occur in the environment. In this paper, a new localization method is proposed to solve the scalability problems, with a new method for detecting and making sense of diverse traffic lines. Like the way human drives, a self-driving system should not rely on an exact position to travel in most scenarios. As a result, without HD Maps, GPS or IMU, the proposed localization subsystem relies only on detecting driving-related features around (like lane lines, stop lines, and merging lane lines). For spotting and reasoning all these features, a new line detector is proposed and tested against multiple datasets.