Multitask learning for large-scale semantic change detection

Change detection is one of the main problems in remote sensing, and is essential to the accurate processing and understanding of the large scale Earth observation data available. Most of the recently proposed change detection methods bring deep learning to this context, but change detection labelled datasets which are openly available are still very scarce, which limits the methods that can be proposed and tested. In this paper we present the first large scale very high resolution semantic change detection dataset, which enables the usage of deep supervised learning methods for semantic change detection with very high resolution images. The dataset contains coregistered RGB image pairs, pixel-wise change information and land cover information. We then propose several supervised learning methods using fully convolutional neural networks to perform semantic change detection. Most notably, we present a network architecture that performs change detection and land cover mapping simultaneously, while using the predicted land cover information to help to predict changes. We also describe a sequential training scheme that allows this network to be trained without setting a hyperparameter that balances different loss functions and achieves the best overall results.