메뉴 건너뛰기
소속 기관 / 학교 인증
인증하면 논문, 학술자료 등을  무료로 열람할 수 있어요.
한국대학교, 누리자동차, 시립도서관 등 나의 기관을 확인해보세요
(국내 대학 90% 이상 구독 중)
고객센터 ENG
주제분류

논문 기본 정보

저자정보
(Brain Ventures) (Brain Ventures) (Brain Ventures) (Brain Ventures)
저널정보
한국지능정보시스템학회 한국지능정보시스템학회 학술대회논문집 2023년 한국지능정보시스템학회 추계학술대회 초록집
오류 신고하기

피인용 0

검색

    초록·키워드

    Object recognition is one of the critical tasks for an autonomous agent. Previous studies mostly dealt with object detection for traffic agents such as cars, bicycles, and buses while the detection of road obstacles remained unatꠓtended. Obstacles on the road must also be detected to avoid collisions and ensure a high level of safety. In this paper, we introduce a novel image dataset with annotation for both categories. We report a preliminary dataset containing 7, 500 images with annotations for 7, 7 classes for the traffic agents and the road obstacles each. We further provide the detection results on a fine-tuned Yolov5 model. Our results suggest that the trained model is capable of running in real-time with a runtime of 17ms and a mean Average Precision of 0.728 on Jetson AGX Orin. In the future, we aim to collect an image dataset from our small mobility platform for better generalization to the real-world environment.

    최근 본 자료 전체보기

      UCI(KEPA) : I410-151-24-02-089213850