基于图神经网络的路面病害态势预测方法
马泽超,刘小明,夏汗青,王伟强,王久增,申海涛

Pavement distress situation prediction method based on graph neural network
Zechao MA,Xiaoming LIU,Hanqing XIA,Weiqiang WANG,Jiuzeng WANG,Haitao SHEN
表 3 多重注意力模块对性能的影响
Tab.3 Effect of multi-attention module on performance
注意力模块MAEMSERMSERecall
仅视图注意力模块5.057083.08539.09820.861
仅时间注意力模块6.0533159.023812.60940.830
全部注意力模块2.610921.14814.59410.933