Preprint
Sep 2026
Ev-YOLO: Uncertainty-Aware Object Detection via a Unified Evidential Formulation
This work proposes an evidential version of YOLOv8 in which both classification and bounding-box regression are formulated within a common evidential framework, allowing the evidential formulation to be applied not only to classification but also to localisation.
Simon Barbarit-Gaboriau, Hind Laghmara, Rémi Boutteau et al.
· 0 citations