Semisupervised Change Detection With Feature-Prediction Alignment

Change detection (CD) has received raising attention for its broad application value. However, traditional fully supervised CD methods have a huge demand for pixel-level annotations, which are laborious and even impossible in some few-shot scenarios. Recently, several semisupervised CD (SSCD) methods have been proposed to utilize numerous unlabeled remote sensing image (RSI) pairs, which can largely reduce the annotation dependence. These methods are mainly based on: 1) adversarial learning, whose optimization direction is difficult to control as a black-box method, or 2) feature-consistency learning, which has no explicit physical meaning. To deal with these difficulties, we propose a novel progressive SSCD framework in this article, termed feature-prediction alignment (FPA). FPA can efficiently utilize unlabeled RSI pairs for training by two alignment strategies. First, a class-aware feature alignment (FA) strategy is designed to align the area-level change/no-change feature extracted from different unlabeled RSI pairs (i.e., across regions) with the awareness of their locations, in order to reduce the feature difference within the same classes. Second, a pixelwise prediction alignment (PA) is devised to align the pixel-level change prediction of strongly augmented unlabeled RSI pairs to the pseudo-labels calculated from the corresponding weakly augmented counterparts, in order to reduce the prediction uncertainty of various RSI transformations with physical meaning. Experiments are carried out on four widely used CD benchmarks, including Learning, Vision and Remote Sensing Laboratory (LEVIR-CD), Wuhan University building CD (WHU-CD), CDD, and GZ-CD, and our FPA achieves the state-of-the-art performance. The experimental results demonstrate the superiority of our method in both effectiveness and generalization. Our code is available at https://github.com/zxt9/FPA-SSCD.