融合前景注意力的轻量级交通标志检测网络

针对目标检测算法模型在交通标志检测上容易出现错检和漏检等问题,提出一种融合前景注意力的轻量级交通标志检测网络YOLOT。首先引入SiLU激活函数,提升模型检测的准确率;其次设计了一种基于鬼影模块的轻量级骨干网络,有效提取目标物特征;接着引入前景注意力感知模块,抑制背景噪声;然后改进路径聚合网络,加入残差结构,充分学习底层特征信息;最后使用VariFocalLoss和GIoU,分别计算目标的分类损失和目标间的相似度,使目标的分类和定位更加准确。在多个数据集上进行了大量实验,结果表明,本文方法的精度优于目前最先进方法,在CCTSDB数据集上进行消融实验,最终精度达到98.50%,与基线模型相比,准确率提升1.32%,同时模型仅4.7 MB,实时检测帧率达到44 FPS。

Lightweight traffic sign detection network with fused foreground attention

A lightweight traffic sign detection network incorporating foreground attention, YOLOT, is proposed to address the problemthat object detection algorithm models are prone to error and miss detection on traffic sign detection. Firstly, the introduction of the SiLUactivation function to improve the accuracy of model detection; secondly, a lightweight backbone network based on the ghost module isdesigned to effectively extract object features; thirdly, introduction of foreground attention perception module to suppress backgroundnoise; fourthly, we improve the path aggregation network by adding a residual structure to the feature fusion process; finally, we useVariFocalLoss and GIoU to calculate the classification loss of objects and the similarity between objects. Extensive experiments areconducted on several datasets, and the results show that the accuracy of the method in this paper is better than the current state-of-the-artmethods. Ablation experiments are conducted on the CCTSDB dataset, and the final accuracy reaches 98. 50%, with an accuracyimprovement of 1. 32% compared to the baseline model, while the model is only 4. 7 MB, and the real-time detection frame rate reaches44 frames per second.