YOLO与专家混合体相遇：面向鲁棒目标检测的自适应专家路由机制

YOLO Meets Mixture-of-Experts: Adaptive Expert Routing for Robust Object Detection

November 17, 2025

作者: Ori Meiraz, Sharon Shalev, Avishai Weizman

cs.AI

摘要

本文提出了一种新颖的混合专家目标检测框架，通过在多路YOLOv9-T专家网络间引入自适应路由机制，实现动态特征 specialization。相较于单一YOLOv9-T模型，该框架在平均精度均值（mAP）和平均召回率（AR）指标上均表现出更优性能。

English

This paper presents a novel Mixture-of-Experts framework for object detection, incorporating adaptive routing among multiple YOLOv9-T experts to enable dynamic feature specialization and achieve higher mean Average Precision (mAP) and Average Recall (AR) compared to a single YOLOv9-T model.