The Change You Want To Detect: Semantic Change Detection In Earth Observation With Hybrid Data Generation

2025-03-19

AI Summary

To address the dual bottlenecks of scarce real-world annotations and limited scene diversity in synthetic data for semantic change detection (SCD) in very-high-resolution (VHR) remote sensing imagery, this paper proposes HySCDG—a semantic- and spatially-guided hybrid synthesis paradigm. HySCDG is the first method to jointly generate realistic VHR imagery and semantically aligned masks via diffusion-based image inpainting. It constructs FSC-180k, a large-scale hybrid training set comprising 180k samples that balances realism and scene diversity. The framework integrates multi-temporal land-cover semantic modeling, hybrid data distillation, and zero-/few-shot transfer learning. Evaluated on five semantic change classes, HySCDG consistently outperforms the fully synthetic baseline SyntheWorld, with substantial gains in pre-trained model performance. All code, models, and datasets are publicly released.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Bi-temporal change detection at scale based on Very High Resolution (VHR) images is crucial for Earth monitoring. This remains poorly addressed so far: methods either require large volumes of annotated data (semantic case), or are limited to restricted datasets (binary set-ups). Most approaches do not exhibit the versatility required for temporal and spatial adaptation: simplicity in architecture design and pretraining on realistic and comprehensive datasets. Synthetic datasets are the key solution but still fail to handle complex and diverse scenes. In this paper, we present HySCDG a generative pipeline for creating a large hybrid semantic change detection dataset that contains both real VHR images and inpainted ones, along with land cover semantic map at both dates and the change map. Being semantically and spatially guided, HySCDG generates realistic images, leading to a comprehensive and hybrid transfer-proof dataset FSC-180k. We evaluate FSC-180k on five change detection cases (binary and semantic), from zero-shot to mixed and sequential training, and also under low data regime training. Experiments demonstrate that pretraining on our hybrid dataset leads to a significant performance boost, outperforming SyntheWorld, a fully synthetic dataset, in every configuration. All codes, models, and data are available here: $href{https://yb23.github.io/projects/cywd/}{https://yb23.github.io/projects/cywd/}$.

Problem

• Addresses semantic change detection in Earth observation using hybrid data generation.
• Overcomes limitations of existing methods requiring large annotated datasets.
• Enhances performance in change detection with a comprehensive hybrid dataset.

Innovation

• Hybrid data generation for semantic change detection
• Combines real and synthetic VHR images effectively
• Enhances performance with comprehensive dataset FSC-180k