-
141
Leveraging Deep Learning and Multimodal Large Language Models for Near-Miss Detection Using Crowdsourced Videos
Published 2025-01-01“…This study underscores the potential of combining deep learning with MLLMs to enhance traffic safety analysis by integrating near-miss data as a key predictive layer. Our approach highlights the importance of leveraging near-miss incidents to proactively enhance road safety, thereby reducing the likelihood of crashes through early intervention and better event understanding.…”
Get full text
Article