Vision-Language Models for Autonomous Driving: CLIP-Based Dynamic Scene Understanding

Elhenawy, Mohammed; Ashqar, Huthaifa I.; Rakotonirainy, Andry; Alhadidi, Taqwa I.; Jaber, Ahmed; Tami, Mohammad Abu

doi:10.3390/electronics14071282

Vision-Language Models for Autonomous Driving: CLIP-Based Dynamic Scene Understanding

Scene understanding is essential for enhancing driver safety, generating human-centric explanations for Automated Vehicle (AV) decisions, and leveraging Artificial Intelligence (AI) for retrospective driving video analysis. This study developed a dynamic scene retrieval system using Contrastive L…