본문 바로가기

Synthetic Training Dataset Generation Using a Digital Twin-based Autonomous Driving Simulator

문서 유형
연구출판물 · 학술지 논문
등재
SCIEScopus
저널 순위
JCR Q4Scopus Q3
저널
SENSORS AND MATERIALS36(9) · pp.4029 · 2024
피인용
Web of Science 0Scopus 5
FWCI (Scopus)
0.51
JCR IF
JCR 2023 1
ISSN
0914-4935

저자

  1. Jang, In-Sung제1저자Elect & Telecommun Res Inst, Mobil Infra Res Sect, 218 Gajeong Ro, Daejeon 34129, South Korea
  2. 이기준Pusan Natl Univ, Sch Comp Sci & Engn, 2 Busandaehak Ro, Busan 46241, South Korea
  3. Joo, Eun-OhDaejeon Univ, Dept Comp Engn, 62 Daehak Ro, Daejeon 34520, South Korea
  4. Kim, Min-Soo교신저자Daejeon Univ, Dept Comp Engn, 62 Daehak Ro, Daejeon 34520, South Korea

초록

Recently, extensive research has been conducted on generating virtual training data in a digital twin-based simulator to reduce the time and cost associated with acquiring high-quality training data necessary for autonomous driving. In this study, we propose an efficient method of generating synthetic training datasets for autonomous driving by combining real-world and virtual training data. Specifically, we propose a method of implementing a digital twin-based autonomous driving simulator, collecting large amounts of virtual training data using its camera sensor, and generating synthetic training datasets by combining virtual and real-world training data in various ratios. The effectiveness of these datasets is then validated in deep learning applications, particularly for detecting traffic lights and signal information. Validation results indicate that synthetic training datasets significantly improve deep learning performance, provided they include a sufficient amount of real-world training data to avoid class imbalance issues. We conclude that the synthetic training datasets generated using a digital twin-based simulator are cost-effective and practical for deep learning applications.

키워드

synthetic training datavirtual training datadeep learningdigital twinautonomous driving

식별자

Scopus EID
2-s2.0-85206088795
WoS UID
001327909600001