基于多维注意力融合的驾驶场景分割增强算法  

Driving scene segmentation enhancement algorithm based on multidimensional attention fusion

作　　者：刘奕晨 章坚武[1] 胡晶 Liu Yichen;Zhang Jianwu;Hu Jing(School of Communication Engineering,Hangzhou Dianzi University,Hangzhou 310018,China;Zhejiang Uniview Technologies Co.,Ltd.,Hangzhou 310051,China)

机构地区：[1]杭州电子科技大学通信工程学院,杭州310018 [2]浙江宇视科技有限公司,杭州310051

出　　处：《计算机应用研究》2023年第10期3180-3185,共6页Application Research of Computers

基　　金：国家自然科学基金资助项目(U1866209,61772162)。

摘　　要：针对使用注意力机制的语义分割模型计算资源消耗与精度不平衡的问题,提出一种轻量化的语义分割注意力增强算法。首先,基于驾驶场景中物体的形状特点设计了条状分维注意力机制,使用条形池化代替传统方形卷积,并结合降维操作分维度提取长程语义关联,削减模型计算量。接着融合通道域与空间域上的注意力,形成可叠加与拆解的轻量化多维注意力融合模块,全方位提取特征信息,进一步提升模型精度。最后,将模块插入基于ResNet-101骨干网的编码—解码网络中,指导高低层语义融合,矫正特征图边缘信息,补充预测细节。实验表明,该模块有较强的鲁棒性和泛化能力,与同类型注意力机制相比,削减了约90%的参数量以及80%的计算量,且分割精度依旧取得了稳定的提升。To address the problem of unbalanced computational resource consumption and accuracy of semantic segmentation models using attention mechanism,this paper proposed a lightweight attention enhancement algorithm for semantic segmentation.Firstly,it designed a striped dimensional attention mechanism based on the shape characteristics of objects in driving scenes,used striped pooling instead of traditional square convolution,and combined dimensionality reduction operations to extract long-range semantic associations in each dimension to cut down the model computation.Then it fused the attention on channel domain and spatial domain to form a lightweight multidimensional attention fusion module that could be superimposed and disassembled to extract feature information in all directions and further improve the model accuracy.Finally,it inserted the module into the ResNet-101 backbone based encoding-decoding network to guide the semantic fusion of high and low layers,correct the feature map edge information,and supplement the prediction details.The experiments show that the module has strong robustness and generalization ability,cutting about 90%of the number of parameters and 80%of the computation compared with the same type of attention mechanism,and the segmentation accuracy still achieves a stable improvement.

关 键 词：语义分割 注意力机制 条状特征提取 多维注意力融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]