To tackle this issue, we propose a novel localization method that extracts reliable semi-dense 2D-3D matching points based on dense keypoint matches.
For data scaling, we introduce a Warmup-Stable-Decay (WSD) learning rate scheduler (LRS), conducive to continuous training and domain adaptation.
Benefiting from the developments in deep learning technology, deep-learning-based algorithms employing automatic feature extraction have achieved remarkable performance on the change detection (CD) task.
Ranked #1 on Change Detection on GoogleGZ-CD