Road network extraction from remotely sensed imagery has become a powerful tool for updating geospatial databases, owing to the success of convolutional neural (CNN) based deep learning semantic segmentation techniques combined with high-resolution that modern remote sensing provides. However, most CNN approaches cannot obtain high precision maps rich details when processing imagery. In this st...

目的 在图像语义分割中，细节特征和语义特征的融合是该领域的一个难点。一些在特定网络架构下设计的专用融合模块缺乏可扩展性和普适性，自注意力虽然可以实现全局的信息捕获，但不能实现不同特征的融合，其他的注意力机制在进行掩码计算时缺少可解释性。本文根据特征图之间的关联度进行建模，提出一种互注意力机制驱动的分割模块。方法 该模块获取不同阶段的细节特征图和语义特征图，建立细节特征图上任一点和语义特征图之间的关联模型，并在关联模型的指导下对语义特征图上的特征进行聚合，作为细节特征图上该特征点的补充，从而将语义特征图上的信息融合到细节特征图上，并进一步采用相同的操作将细节特征图上的信息融合到语义特征图上，实现来自不同阶段特征图的相互融合。结果 选取5个语义分割模型进行实验，实验结果表明，在使用替换方式对BiSeNet V2 （bilateral segmentation network）进行修改之后...

With the increasing demand of autonomous systems, pixelwise semantic segmentation for visual scene understanding needs to be not only accurate but also efficient potential real-time applications. In this paper, we propose Context Aggregation Network, a dual branch convolutional neural network, with significantly lower computational costs as compared state-of-the-art, while maintaining competiti...

Objectives To formalize the relations between the different Semantic Types in the Semantic Network of the Unified Medical Language System (UMLS) in the context of computer interpretable task-based clinical guidelines model. Design We used the Semantic Type Collections as our basis in the formalization. We defined some operators based on the relations which we considered were applicable to all t...