基于多模态知识对齐的双分支点云语义分割算法
杨军,卯恒睿,党吉圣

Dual-branch point cloud semantic segmentation algorithm based on multimodal knowledge alignment
Jun YANG,Hengrui MAO,Jisheng DANG
表 2 NuScenes数据集上不同算法的语义分割性能对比
Tab.2 Comparison of semantic segmentation performance across different methods on NuScenes dataset
模型mIoU/%IoU/%
护栏自行车公交车汽车施工区域摩托车行人交通锥
Cylinder3D[18]77.282.829.884.389.463.079.377.273.4
PMF[25]77.082.040.081.088.064.079.080.076.0
RViT-FLARES[22]77.076.739.293.092.055.281.677.264.9
PC-BEV[23]78.878.246.392.593.455.087.181.065.4
2DPASS[24]78.576.152.695.785.156.591.780.968.2
本文模型79.376.552.396.188.256.691.580.669.2
模型IoU/%
拖车卡车可行驶区域其他地面人行道地形人造结构植被
Cylinder3D[18]84.669.197.770.280.375.590.487.6
PMF[25]81.067.097.068.078.074.090.088.0
RViT-FLARES[22]70.984.196.874.195.675.188.686.7
PC-BEV[23]69.285.797.176.877.076.390.688.5
2DPASS[24]74.682.296.676.472.471.588.287.3
本文模型76.588.696.575.273.272.388.087.5