Attention-augmented Spatio-Temporal Segmentation for Land Cover Mapping

TL;DR

This paper presents a novel neural network architecture combining UNet, Bidirectional LSTM, and Attention mechanisms to improve land cover mapping from satellite data by exploiting spatial and temporal features.

Contribution

The proposed model effectively integrates spatial and temporal information with attention to enhance land cover classification accuracy, addressing noise and class-specific temporal patterns.

Findings

Outperforms state-of-the-art methods on real-world datasets.

Effectively mitigates noise and identifies key temporal periods.

Visualizations show attention weights focus on discriminative time frames.

Abstract

The availability of massive earth observing satellite data provide huge opportunities for land use and land cover mapping. However, such mapping effort is challenging due to the existence of various land cover classes, noisy data, and the lack of proper labels. Also, each land cover class typically has its own unique temporal pattern and can be identified only during certain periods. In this article, we introduce a novel architecture that incorporates the UNet structure with Bidirectional LSTM and Attention mechanism to jointly exploit the spatial and temporal nature of satellite data and to better identify the unique temporal patterns of each land cover. We evaluate this method for mapping crops in multiple regions over the world. We compare our method with other state-of-the-art methods both quantitatively and qualitatively on two real-world datasets which involve multiple land cover classes. We also visualise the attention weights to study its effectiveness in mitigating noise and identifying discriminative time period.