At the 2023 Seoul Mobility Show(https://lnkd.in/gVxGSU9S) , we exhibited the multi-agent metaverse platform on behalf of KAIST Cho Chun Sik Mobility Graduate School. The KAIST campus was made into a virtual environment for the children who visited our booth and the experience of exploring the campus by walking, running, and driving in the virtual environment.
The self-driving car developed by Prof. Dongsuk Kum(http://vdclab.kaist.ac.kr/) in our department used the campus as a test bed and trained as a safer vehicle. This is physically time-consuming and costly. Thousands of scenarios can be tested without limit by building a virtual environment, and situations that are very difficult in real life can be created in the virtual environment. These abnormal driving behaviors become very important input data for autonomous vehicles.
Our laboratory will continue to make efforts to create safer roads.
https://lnkd.in/gyCNv97Q

86회 춘계 교통학회 학술대회에 8편의 논문을 발표하였습니다. 

• Road Accident prediction in Daejeon City Based on Deep Learning
• 시계열 데이터 증강을 통한 교통량 예측 성능 향상 연구
• 대전시 교통 데이터 웨어하우스 시스템 고도화 연구
• 단일교차로의 DQN 기반 강화학습의 최적 보상 방안 연구
• 데이터 분석용 플랫폼을 이용한 보행자 교통사고 분석
• 보행자 및 자전거 전용도로를 고려한 신호교차로 용량분석
• 자율주행을 고려한 교차로 용량 분석
• 초소형 전기차: 문헌 고찰을 통한 국내외 연구 동향 분석