Comparative Analysis of Deep Neural Networks YOLOv11 and YOLOv12 for Real-Time Vehicle Detection in Autonomous Vehicles
Résumé
Accurate, real-time vehicle detection is crucial for autonomous vehicles navigating dynamic traffic environments.This study compares YOLOv11 and the newly released YOLOv12, two state-of-the-art deep learning models for object detection, to assess enhancements in speed, accuracy, and robustness.YOLOv12 has improved upon YOLOv11's architecture with an attention mechanism and Residual Efficient Layer Aggregation Networks (R-ELAN).The improvements for YOLOv12 are designed to obtain better accuracy and improved computational performance as compared to YOLOv11.YOLOv11 and YOLOv12 were trained and tested on a newly developed dataset with 38,500 fully annotated images of seven classes of vehicles taken in different environmental conditions.Results show YOLOv12 achieves higher recall (95.0%),F1score (96.03%), and mAP@50-95 (88.6%), while both maintain real-time inference speeds.YOLOv12 also demonstrated an improved capacity to detect small or partially occluded objects in challenging scenes.Overall, these findings establish YOLOv12 as a better solution for perceiving real-time data while autonomous driving, with a real prospect for implementation in intelligent transportation systems and edge-computing.
Citer ce document
Exporter : BibTeX · RIS (Zotero, Mendeley, EndNote)
Accès au document
Texte intégral archivé sur SPHAERO, en accès ouvert : lisible gratuitement en ligne après connexion. Le document reste aussi disponible à sa source. Se connecter
Voir l'article sur le site de la revueLicence et provenance
Licence : CC BY
Notice moissonnée depuis OpenAlex le 28/09/2026. Le document reste hébergé par sa source.
Voir le document à la source →