OPTIMUS: Observing Persistent Transformations in Multi-Temporal Unlabeled Satellite-Data

2025-02-26

IEEE Workshop/Winter Conference on Applications of Computer Vision

AI Summary

Addressing the scarcity of annotated rare persistent land-cover change instances in remote sensing imagery, this paper proposes a fully self-supervised approach that eliminates reliance on manual annotations. The core methodological innovation lies in leveraging the temporal ordering of satellite image sequences as an implicit proxy signal for persistent change—first introduced in this work—and learning change-sensitive representations via contrastive temporal representation learning, coupled with change-point detection (CPD) for weakly supervised localization. This framework enables end-to-end, annotation-free persistent change identification without requiring labeled rare-change samples. Evaluated on standard benchmarks, our method achieves an AUROC of 87.6%, outperforming prior baselines by 31.3 percentage points. To foster reproducibility and future research, we publicly release both the source code and a newly curated benchmark dataset specifically designed for rare persistent change detection.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Unsupervised & Self-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

In the face of pressing environmental issues in the 21st century, monitoring surface changes on Earth is more important than ever. Large-scale remote sensing, such as satellite imagery, is an important tool for this task. However, using supervised methods to detect changes is difficult because of the lack of satellite data annotated with change labels, especially for rare categories of change. Annotation proves challenging due to the sparse occurrence of changes in satellite images. Even within a vast collection of images, only a small fraction may exhibit persistent changes of interest. To address this challenge, we introduce OPTIMUS, a self-supervised learning method based on an intuitive principle: if a model can recover information about the relative order of images in the time series, then that implies that there are long-lasting changes in the images. OPTIMUS demonstrates this principle by using change point detection methods on model outputs in a time series. We demonstrate that OPTIMUS can directly detect interesting changes in satellite images, achieving an improvement in AUROC score from 56.3% to 87.6% at distinguishing changed time series from unchanged ones compared to baselines. Our code and dataset are available at https://huggingface.co/datasets/optimus-change/optimus-dataset/.

Problem

• Detecting persistent changes in unlabeled satellite time-series data
• Overcoming lack of annotated change labels in remote sensing
• Improving rare category change detection via self-supervised learning

Innovation

• Self-supervised learning for satellite data
• Change point detection in time series
• Relative order recovery for persistent changes