[论文] Traffic Sign Recognition for Autonomous Driving Using Branched YOLOv2 ...

论文概要

研究领域: CV
作者: Arefeh Rezaei
发布时间: 2026-09-18
arXiv: 2609.22060

中文摘要

交通标志识别（TSR）是自动驾驶和高级驾驶辅助系统的重要感知任务，系统须同时定位交通标志并判定其语义类别。本文提出一个基于 YOLOv2 的 TSR 系统，用于同时完成检测与分类，并研究两项互补改进。第一，为 YOLOv2 增加中间预测层，构成分支架构，可在简单样本上提前终止推理、降低计算时间，并研究了整图级与单元级两种分支策略。第二，引入几何信息以减少视觉相似标志间的分类错误：一种无监督贝叶斯图像分割方法生成二值表示，在 YOLOv2 的边界框内与类别特定的几何模板比对；该信息既可在推理阶段使用，也可作为训练期间的附加信号。我们通过无缝克隆和受控图像变换组合 GTSDB 与 GTSRB 样本构建了专门数据集。实验覆盖十个交通标志类别，含 3,000 个训练样本和 300 个测试样本。选定的分支架构运行时间为 0.647 秒、mAP 为 0.680，而基线 YOLOv2 为 0.6607 秒、0.680 mAP。推理阶段加入几何验证将 mAP 提升至 0.713；几何特征训练变体取得 0.697 mAP，运行时间 0.6608 秒。

原文摘要

Traffic sign recognition (TSR) is an important perception task for autonomous driving and advanced driver-assistance systems, where a system must both localize traffic signs and determine their semantic classes efficiently. This work presents a TSR system based on YOLOv2 for simultaneous detection and classification. Two complementary modifications are studied. First, YOLOv2 is extended with intermediate prediction layers, forming a branched architecture that can terminate inference early for easy cases and reduce computation time. Both whole-image and cell-wise branching strategies are investigated. Second, geometric information is introduced to reduce classification errors between visually similar signs. An unsupervised Bayesian image-segmentation method produces binary representations t...

自动采集于 2026-09-22

#论文 #arXiv #CV #小凯