# ChangeDiff: A Multi-Temporal Change Detection Data Generator with Flexible Text Prompts via Diffusion Model

arXiv: https://arxiv.org/abs/2412.15541 · HTML: https://mlsanity.com/paper/2412.15541/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- Placed in Inverse problems and image restoration, one of 14 directions in Diffusion models.
- Builds on 6 of its 26 references that are in the map.
- First posted on arXiv 20 Dec 2024.

## Placed in

- Diffusion models / [Inverse problems and image restoration](https://mlsanity.com/diffusion/inverse-problems-and-image-restoration/)

## Builds on (most cited first, 6 of 6)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation](https://mlsanity.com/paper/2110.08733/) (2021)
- [DiffusionSat: A Generative Foundation Model for Satellite Imagery](https://mlsanity.com/paper/2312.03606/) (2023)
- [Changen2: Multi-Temporal Remote Sensing Generative Change Foundation Model](https://mlsanity.com/paper/2406.17998/) (2024)

## Built on by (most cited first, 10 of 10)

- [FisherTune: Fisher-Guided Robust Tuning of Vision Foundation Models for Domain Generalized Segmentation](https://mlsanity.com/paper/2503.17940/) (2025)
- [TerraGen: A Unified Multi-Task Layout Generation Framework for Remote Sensing Data Augmentation](https://mlsanity.com/paper/2510.21391/) (2025)
- [Open-Vocabulary Domain Generalization in Urban-Scene Segmentation](https://mlsanity.com/paper/2602.18853/) (2026)
- [Beyond Quadratic: Linear-Time Change Detection with RWKV](https://mlsanity.com/paper/2603.19606/) (2026)
- [ChangeBridge: Spatiotemporal Image Generation with Multimodal Controls for Remote Sensing](https://mlsanity.com/paper/2507.04678/) (2025)
- [OSMGen: Highly Controllable Satellite Image Synthesis using OpenStreetMap Data](https://mlsanity.com/paper/2511.00345/) (2025)
- [JL1-CC&QA: Extending the JL1-CD Benchmark with Change Captioning and Question Answering](https://mlsanity.com/paper/2606.31745/) (2026)
- [Learning Semantic-Robust Change Detection via Semantic-Invariant Self-Distillation](https://mlsanity.com/paper/2607.19000/) (2026)
- [FlowForm: Synergizing Fluid Physics with Topological Consistency for Satellite Flood Synthesis](https://mlsanity.com/paper/2608.03822/) (2026)
- [Real-World Knowledge-Guided Change Data Synthesis for Remote Sensing](https://mlsanity.com/paper/2608.24263/) (2026)

## Abstract (arXiv)

Data-driven deep learning models have enabled tremendous progress in change detection (CD) with the support of pixel-level annotations. However, collecting diverse data and manually annotating them is costly, laborious, and knowledge-intensive. Existing generative methods for CD data synthesis show competitive potential in addressing this issue but still face the following limitations: 1) difficulty in flexibly controlling change events, 2) dependence on additional data to train the data generators, 3) focus on specific change detection tasks. To this end, this paper focuses on the semantic CD (SCD) task and develops a multi-temporal SCD data generator ChangeDiff by exploring powerful diffusion models. ChangeDiff innovatively generates change data in two steps: first, it uses text prompts and a text-to-layout (T2L) model to create continuous layouts, and then it employs layout-to-image (L2I) to convert these layouts into images. Specifically, we propose multi-class distribution-guided text prompts (MCDG-TP), allowing for layouts to be generated flexibly through controllable classes and their corresponding ratios. Subsequently, to generalize the T2L model to the proposed MCDG-TP, a class distribution refinement loss is further designed as training supervision. %For the former, a multi-classdistribution-guided text prompt (MCDG-TP) is proposed to complement via controllable classes and ratios. To generalize the text-to-image diffusion model to the proposed MCDG-TP, a class distribution refinement loss is designed as training supervision. For the latter, MCDG-TP in three modes is proposed to synthesize new layout masks from various texts. Our generated data shows significant progress in temporal continuity, spatial diversity, and quality realism, empowering change detectors with accuracy and transferability. The code is available at https://github.com/DZhaoXd/ChangeDiff