Image augmentation with invertible networks in interactive satellite image change detection

2025-10-21

AI Summary

To address the high annotation cost and limited efficacy of conventional data augmentation in satellite image change detection, this paper proposes an interactive detection framework integrating active learning with invertible neural networks. An iterative query mechanism selects the most informative samples for human annotation, while an invertible network maps inputs into a linear latent space where differentiable and invertible augmentations are performed—enabling efficient sample expansion and dynamic model updating. The key contribution is the first incorporation of invertible networks into the latent-space augmentation paradigm for change detection, ensuring transformation fidelity and interpretability. Experiments on multiple benchmark datasets demonstrate significant improvements in detection accuracy and convergence speed, achieving state-of-the-art performance with substantially fewer annotated samples.

Technology Category

• Machine Learning: Active Learning
• Computer Vision: Diffusion Models for Vision
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

This paper devises a novel interactive satellite image change detection algorithm based on active learning. Our framework employs an iterative process that leverages a question-and-answer model. This model queries the oracle (user) about the labels of a small subset of images (dubbed as display), and based on the oracle's responses, change detection model is dynamically updated. The main contribution of our framework resides in a novel invertible network that allows augmenting displays, by mapping them from highly nonlinear input spaces to latent ones, where augmentation transformations become linear and more tractable. The resulting augmented data are afterwards mapped back to the input space, and used to retrain more effective change detection criteria in the subsequent iterations of active learning. Experimental results demonstrate superior performance of our proposed method compared to the related work.

Problem

• Augmenting satellite images using invertible networks for change detection
• Developing interactive active learning framework with user queries
• Mapping nonlinear image transformations to linear latent spaces

Innovation

• Invertible networks enable nonlinear-to-linear data mapping
• Active learning queries users for iterative label refinement
• Latent space augmentation improves change detection accuracy