ChangeJEPA improves remote sensing change detection accuracy by predicting states
ChangeJEPA: Predict Before You Compare for Remote Sensing Change Detection
Computer Vision and Pattern Recognition
Summary
Detecting changes in satellite or aerial images is tricky because differences can come from lighting, seasons, or camera angles, not just real changes on the ground. The authors designed ChangeJEPA, a method that predicts how the scene should look after a change using both the old image and part of the new image. By focusing on predicting the underlying scene instead of just comparing images directly, ChangeJEPA better tells apart real changes from irrelevant variations. Tests on multiple datasets showed that it improves accuracy compared to older methods.
What this means in practice
- •For urban planning teams: Improve detection of meaningful land use changes in satellite imagery despite seasonal and lighting variations.
- •For environmental monitoring groups: Enhance tracking of deforestation or natural disaster impacts by predicting latent scene changes rather than relying only on direct image comparisons.
Authors
Buddhi Wijenayake, Dhananjaya Jayasundara, Vishal M. Patel
Abstract
Remote sensing change detection is typically formulated as Siamese feature comparison, where two images are encoded independently and a decoder determines which discrepancies correspond to meaningful change. This is challenging because the encoders preserve both true scene changes and nuisance variations such as illumination, seasonal, radiometric, and geometric differences. We introduce ChangeJEPA, a bitemporal joint-embedding predictive architecture that reformulates change detection as latent post-state prediction. Given a pre-change image and partial post-change evidence, ChangeJEPA predicts the full post-change representation under two complementary conditions: a prior hypothesis derived from the pre-change scene and an updated hypothesis conditioned on observed post-change context. A joint-embedding predictive loss aligns both hypotheses with the target latent representation, encouraging semantic consistency while suppressing transient appearance variations. Change is then inferred from the disagreement among the pre-change representation, the observed post-change target, and the predicted latent states, capturing complementary evidence of change, prior surprise, contextual update, and prediction reliability. Across SYSU-CD, LEVIR-CD+, and CDD, ChangeJEPA consistently outperforms its decoder matched Siamese baseline by up to 1.51 IoU points and achieves the best IoU and F1-score on all three benchmarks, demonstrating the advantage of latent post-state prediction over direct feature comparison. Page: https://buddhi19.github.io/ChangeJEPA