Improving Zero-Shot Object-Level Change Detection by Incorporating Visual Correspondence

2025-01-09

AI Summary

Existing methods for zero-shot object-level change detection in cross-view image pairs suffer from three key limitations: (1) high false positive rates on unchanged samples, (2) absence of explicit change localization—i.e., lack of pixel-wise correspondence between pre- and post-change regions, and (3) poor cross-domain generalization. To address these, this work introduces *change correspondence modeling* into an end-to-end trainable framework—the first such approach. We propose a novel correspondence prediction mechanism that jointly leverages homography estimation and Hungarian matching, augmented with explicit change-location supervision to improve spatial precision. Evaluated on both in-distribution and zero-shot cross-domain benchmarks, our method achieves state-of-the-art performance: it significantly reduces false positives while simultaneously advancing both change detection accuracy and pixel-level correspondence localization.

Technology Category

• Computer Vision: Object Detection & Categorization
• Machine Learning: Multi-instance/Multi-view Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Detecting object-level changes between two images across possibly different views is a core task in many applications that involve visual inspection or camera surveillance. Existing change-detection approaches suffer from three major limitations: (1) lack of evaluation on image pairs that contain no changes, leading to unreported false positive rates; (2) lack of correspondences (ie, localizing the regions before and after a change); and (3) poor zero-shot generalization across different domains. To address these issues, we introduce a novel method that leverages change correspondences (a) during training to improve change detection accuracy, and (b) at test time, to minimize false positives. That is, we harness the supervision labels of where an object is added or removed to supervise change detectors, improving their accuracy over previous work by a large margin. Our work is also the first to predict correspondences between pairs of detected changes using estimated homography and the Hungarian algorithm. Our model demonstrates superior performance over existing methods, achieving state-of-the-art results in change detection and change correspondence accuracy across both in-distribution and zero-shot benchmarks.

Problem

• Image Comparison
• Change Detection
• Environmental Robustness

Innovation

• Change Detection
• Correspondence Algorithm
• Improved Accuracy