Search papers, labs, and topics across Lattice.
The authors introduce DenseUIS, a new high-resolution remote sensing dataset tailored for building and road extraction in dense urban informal settlements, comprising imagery from 126 urban villages in Shenzhen and Guangzhou, China. They benchmarked state-of-the-art deep learning models on DenseUIS, finding that existing methods struggle with the unique morphological characteristics of these environments. This highlights the need for specialized models and provides a challenging benchmark for future research in this domain.
Existing deep learning models falter when mapping buildings and roads in dense urban informal settlements, revealing a critical gap in remote sensing capabilities for these challenging environments.
As a widespread form of informal settlements, urban villages present significant challenges for sustainable urban development and governance. Precise mapping of their infrastructure is essential, however, existing remote sensing datasets primarily focus on formal urban environments, lacking fine-grained annotated data for the high-density building patterns and narrow road networks typical of urban villages. To address this gap, we introduce the \textit{DenseUIS} dataset, the first high-resolution remote sensing dataset specifically designed for building and road extraction in extremely dense urban informal settlements, covering 126 urban villages across Shenzhen and Guangzhou in China. Furthermore, we conduct a comprehensive evaluation of state-of-the-art deep learning models on this dataset. Experimental results reveal the limitations of existing methods in handling the unique morphological patterns of dense informal settlements, underscoring the need for specialized approaches. \textit{DenseUIS} therefore provides a robust benchmark for advancing fine-grained urban mapping in complex and high-density informal environments. The dataset is publicly available at https://github.com/rui-research/DenseUIS.
