LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation

TL;DR

LoveDA addresses generalization gaps in remote sensing land-cover segmentation by introducing a two-domain, urban-versus-rural dataset suitable for semantic segmentation and unsupervised domain adaptation. It benchmarks a range of segmentation and UDA methods, analyzes challenges from multi-scale objects and complex backgrounds to inconsistent class distributions, and finds self-training approaches often outperform adversarial ones on cross-domain tasks. The work provides a valuable resource for developing transferable, scalable land-cover mapping models applicable to diverse geographies. Overall, LoveDA motivates improved domain-robust representations for high-resolution remote sensing imagery and large-scale environmental monitoring.

Abstract

Deep learning approaches have shown promising results in remote sensing high spatial resolution (HSR) land-cover mapping. However, urban and rural scenes can show completely different geographical landscapes, and the inadequate generalizability of these algorithms hinders city-level or national-level mapping. Most of the existing HSR land-cover datasets mainly promote the research of learning semantic representation, thereby ignoring the model transferability. In this paper, we introduce the Land-cOVEr Domain Adaptive semantic segmentation (LoveDA) dataset to advance semantic and transferable learning. The LoveDA dataset contains 5987 HSR images with 166768 annotated objects from three different cities. Compared to the existing datasets, the LoveDA dataset encompasses two domains (urban and rural), which brings considerable challenges due to the: 1) multi-scale objects; 2) complex background samples; and 3) inconsistent class distributions. The LoveDA dataset is suitable for both land-cover semantic segmentation and unsupervised domain adaptation (UDA) tasks. Accordingly, we benchmarked the LoveDA dataset on eleven semantic segmentation methods and eight UDA methods. Some exploratory studies including multi-scale architectures and strategies, additional background supervision, and pseudo-label analysis were also carried out to address these challenges. The code and data are available at https://github.com/Junjue-Wang/LoveDA.

Figures (10)

• Figure 1: Overview of the dataset distribution. The images were collected from Nanjing, Changzhou, and Wuhan cities, covering 18 different administrative districts.
• Figure 2: Statistics for the pixels and objects in LoveDA dataset. (a) Number of objects vs. number of pixels. The radius of the circles represents the number of classes. (b) Histogram of the number of objects for each class. (c) Histogram of the number of pixels for each class.
• Figure 3: Statistics for the urban and rural scenes in Nanjing City. (a) Class distribution. (b) Spectral statistics. The mean and standard deviation ($\sigma$) for 5 urban and 5 rural areas are reported. (c) Distribution of the building sizes. The Jianye (urban) and Lishui (rural) scenes are reported.
• Figure 4: Representative confusion matrices for the semantic segmentation experiments.
• Figure 5: Semantic segmentation results on images from the LoveDA Test set in the Liuhe (Rural) area. Some small-scale objects such as buildings and scattered trees are hard to recognize. The forest and agricultural classes are easy to misclassify due to their similar spectra.