Geography-Aware Self-Supervised Learning
This work is significant for researchers and practitioners working with geo-located datasets, particularly in remote sensing, by improving the effectiveness of self-supervised learning when labeled data is limited.
This paper addresses the performance gap between contrastive and supervised learning on geo-located datasets like remote sensing, where unlabeled data is plentiful but labeled data is scarce. The authors propose novel training methods that leverage the spatio-temporal structure of remote sensing data, such as using spatially aligned images over time to create temporal positive pairs and geo-location for pre-text tasks. Their method successfully closes the performance gap across image classification, object detection, and semantic segmentation tasks in remote sensing, and also improves performance on geo-tagged ImageNet images.
Contrastive learning methods have significantly narrowed the gap between supervised and unsupervised learning on computer vision tasks. In this paper, we explore their application to geo-located datasets, e.g. remote sensing, where unlabeled data is often abundant but labeled data is scarce. We first show that due to their different characteristics, a non-trivial gap persists between contrastive and supervised learning on standard benchmarks. To close the gap, we propose novel training methods that exploit the spatio-temporal structure of remote sensing data. We leverage spatially aligned images over time to construct temporal positive pairs in contrastive learning and geo-location to design pre-text tasks. Our experiments show that our proposed method closes the gap between contrastive and supervised learning on image classification, object detection and semantic segmentation for remote sensing. Moreover, we demonstrate that the proposed method can also be applied to geo-tagged ImageNet images, improving downstream performance on various tasks. Project Webpage can be found at this link geography-aware-ssl.github.io.