Meta-Learning for Few-Shot Land Cover Classification
This work addresses the challenge of land cover classification in Earth sciences where data diversity across regions requires large datasets, offering a few-shot adaptation approach that could reduce data needs for domain-specific applications.
The authors tackled the problem of land cover classification across diverse geographic regions by framing it as a few-shot inductive transfer learning task, where a model adapts to unseen regions with minimal data. They found that model-agnostic meta-learning (MAML) outperformed traditional pre-training and fine-tuning on datasets like Sen12MS and DeepGlobe when source and target domains differ, indicating its benefit for Earth science tasks with high regional diversity.
The representations of the Earth's surface vary from one geographic region to another. For instance, the appearance of urban areas differs between continents, and seasonality influences the appearance of vegetation. To capture the diversity within a single category, like as urban or vegetation, requires a large model capacity and, consequently, large datasets. In this work, we propose a different perspective and view this diversity as an inductive transfer learning problem where few data samples from one region allow a model to adapt to an unseen region. We evaluate the model-agnostic meta-learning (MAML) algorithm on classification and segmentation tasks using globally and regionally distributed datasets. We find that few-shot model adaptation outperforms pre-training with regular gradient descent and fine-tuning on (1) the Sen12MS dataset and (2) DeepGlobe data when the source domain and target domain differ. This indicates that model optimization with meta-learning may benefit tasks in the Earth sciences whose data show a high degree of diversity from region to region, while traditional gradient-based supervised learning remains suitable in the absence of a feature or label shift.