arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.09064cs.CVeess.IV
收藏

用于建筑物远程提取的 U-Net、ResUnet 和 U-Net3+ 中的双重跳跃连接

Dual skip connections in U-Net, ResUnet and U-Net3+ for remote extraction of buildings

The University of Melbourne(墨尔本大学)

Bipul Neupane, Jagannath Aryal, Abbas Rajabifard

AI总结:

本文为 U-Net、ResUnet 和 U-Net3+ 设计选择性双重跳跃连接,在多分辨率建筑物数据集上以更少参数取得最高 0.905 F1。

摘要:

城市建筑物通过使用诸如 U-Net 及其后继模型之类的语义分割网络,从高分辨率地球观测(EO)图像中被提取出来。每次迭代都旨在通过采用更密集的跳跃连接机制来提高性能,该机制利用多尺度特征以实现精确的对象制图。然而,更密集的连接会增加网络参数,且不一定有助于精确分割。本文中,我们为三个网络(U-Net、ResUnet 和 U-Net3+)开发了三种双重跳跃连接机制,以选择性地加深必要的特征图,从而提升性能。这三种机制在不同尺度的特征图上进行评估,产生了九种新的网络配置。我们在四个具有不同空间分辨率的建筑物足迹数据集上将它们与其原始标准配置进行比较,其中包括我们为复杂城市环境开发的一个多分辨率(0.3+0.6+1.2 米)数据集。评估表明,在 U-Net 和 U-Net3+ 中密集化大尺度和小尺度特征可产生高达 0.905 的 F1,在我们的新数据集上高于 TransUnet(0.903)和 Swin-Unet(0.882),同时参数最多减少 19 倍。结果表明,选择性地密集化特征图和跳跃连接可在不显著增加参数的情况下增强网络性能。这些发现和新数据集将有助于计算机视觉领域和城市规划决策过程。

英文摘要:

Urban buildings are extracted from high-resolution Earth observation (EO) images using semantic segmentation networks like U-Net and its successors. Each re-iteration aims to improve performance by employing a denser skip connection mechanism that harnesses multi-scale features for accurate object mapping. However, denser connections increase network parameters and do not necessarily contribute to precise segmentation. In this paper, we develop three dual skip connection mechanisms for three networks (U-Net, ResUnet, and U-Net3+) to selectively deepen the essential feature maps for improved performance. The three mechanisms are evaluated on feature maps of different scales, producing nine new network configurations. They are evaluated against their original vanilla configurations on four building footprint datasets of different spatial resolutions, including a multi-resolution (0.3+0.6+1.2m) dataset that we develop for complex urban environments. The evaluation revealed that densifying the large- and small-scale features in U-Net and U-Net3+ produce up to 0.905 F1, more than TransUnet (0.903) and Swin-Unet (0.882) in our new dataset with up to 19x fewer parameters. The results conclude that selectively densifying feature maps and skip connections enhances network performance without a substantial increase in parameters. The findings and the new dataset will contribute to the computer vision domain and urban planning decision processes.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑