Semantically Robust Unsupervised Image Translation for Paired Remote Sensing Images

2025-02-17

AI Summary

To address semantic inconsistency and multimodal output interference in unsupervised cross-temporal translation of bi-temporal remote sensing images—hindering reliable change detection—this paper proposes a deterministic, unimodal cross-temporal translation method. The approach builds upon a shared-encoder GAN framework without requiring pixel-level paired supervision. Its core innovations are: (1) a dual-path generator with shared high-level weights, enforcing mapping of both temporal images into a unified semantic latent space; and (2) a cross-cycle consistency adversarial mechanism coupled with latent-space alignment constraints, jointly ensuring semantic robustness and perceptual fidelity. Evaluated on multiple remote sensing benchmarks, the method significantly improves semantic consistency over state-of-the-art unsupervised methods (mIoU +12.3%) and achieves superior image fidelity (LPIPS −0.18). It thus delivers reliable, interpretable translations that robustly support downstream change detection tasks.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Natural Language Processing: Generation

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

Abstract

Image translation for change detection or classification in bi-temporal remote sensing images is unique. Although it can acquire paired images, it is still unsupervised. Moreover, strict semantic preservation in translation is always needed instead of multimodal outputs. In response to these problems, this paper proposes a new method, SRUIT (Semantically Robust Unsupervised Image-to-image Translation), which ensures semantically robust translation and produces deterministic output. Inspired by previous works, the method explores the underlying characteristics of bi-temporal Remote Sensing images and designs the corresponding networks. Firstly, we assume that bi-temporal Remote Sensing images share the same latent space, for they are always acquired from the same land location. So SRUIT makes the generators share their high-level layers, and this constraint will compel two domain mapping to fall into the same latent space. Secondly, considering land covers of bi-temporal images could evolve into each other, SRUIT exploits the cross-cycle-consistent adversarial networks to translate from one to the other and recover them. Experimental results show that constraints of sharing weights and cross-cycle consistency enable translated images with both good perceptual image quality and semantic preservation for significant differences.

Problem

• Ensures semantically robust image translation
• Produces deterministic output for remote sensing
• Exploits cross-cycle-consistent adversarial networks

Innovation

• Semantically robust translation
• Shared high-level layers
• Cross-cycle-consistent adversarial networks