# Deep Image Translation with an Affinity-Based Change Prior for Unsupervised Multimodal Change Detection

arXiv: https://arxiv.org/abs/2001.04271 · HTML: https://mlsanity.com/paper/2001.04271/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 13 Jan 2020.

## Built on by (most cited first, 22 of 22)

- [Bi-Temporal Semantic Reasoning for the Semantic Change Detection in HR Remote Sensing Images](https://mlsanity.com/paper/2108.06103/) (2021)
- [Change Detection Methods for Remote Sensing in the Last Decade: A Comprehensive Review](https://mlsanity.com/paper/2305.05813/) (2023)
- [Exchange means change: an unsupervised single-temporal change detection framework based on intra- and inter-image patch exchange](https://mlsanity.com/paper/2310.00689/) (2023)
- [Unsupervised Multimodal Change Detection Based on Structural Relationship Graph Representation Learning](https://mlsanity.com/paper/2210.00941/) (2022)
- [Null-text Guidance in Diffusion Models is Secretly a Cartoon-style Creator](https://mlsanity.com/paper/2305.06710/) (2023)
- [A Survey of Sample-Efficient Deep Learning for Change Detection in Remote Sensing: Tasks, Strategies, and Challenges](https://mlsanity.com/paper/2502.02835/) (2025)
- [Federated CycleGAN for Privacy-Preserving Image-to-Image Translation](https://mlsanity.com/paper/2106.09246/) (2021)
- [UCDFormer: Unsupervised Change Detection Using a Transformer-driven Image Translation](https://mlsanity.com/paper/2308.01146/) (2023)
- [Graph Signal Processing for Heterogeneous Change Detection Part I: Vertex Domain Filtering](https://mlsanity.com/paper/2208.01881/) (2022)
- [Transformer-based Multimodal Change Detection with Multitask Consistency Constraints](https://mlsanity.com/paper/2310.09276/) (2023)
- [NN-Copula-CD: A Copula-Guided Interpretable Neural Network for Change Detection in Heterogeneous Remote Sensing Images](https://mlsanity.com/paper/2303.17448/) (2023)
- [OSDM-MReg: Multimodal Image Registration based One Step Diffusion Model](https://mlsanity.com/paper/2504.06027/) (2025)
- [Segmenting France Across Four Centuries](https://mlsanity.com/paper/2505.24824/) (2025)
- [On the potential of sequential and non-sequential regression models for Sentinel-1-based biomass prediction in Tanzanian miombo forests](https://mlsanity.com/paper/2106.15020/) (2021)
- [Recognition of polar lows in Sentinel-1 SAR images with deep learning](https://mlsanity.com/paper/2203.16401/) (2022)
- [Lightweight Change Detection in Heterogeneous Remote Sensing Images with Online All-Integer Pruning Training](https://mlsanity.com/paper/2405.01920/) (2024)
- [Cross-Domain Separable Translation Network for Multimodal Image Change Detection](https://mlsanity.com/paper/2407.16158/) (2024)
- [Graph Signal Processing for Heterogeneous Change Detection Part II: Spectral Domain Analysis](https://mlsanity.com/paper/2208.01905/) (2022)
- [Monitoring snow avalanches from SAR data with deep learning](https://mlsanity.com/paper/2502.18157/) (2025)
- [Deep Diffusion Models and Unsupervised Hyperspectral Unmixing for Realistic Abundance Map Synthesis](https://mlsanity.com/paper/2506.13484/) (2025)
- [Synthetic Aperture Radar Image Change Detection Based on Global Dynamic Context-Aware Network](https://mlsanity.com/paper/2605.16764/) (2026)
- [Content-Induced Spatial-Spectral Aggregation Network for Change Detection in Remote Sensing Images](https://mlsanity.com/paper/2606.10328/) (2026)

## Abstract (arXiv)

Image translation with convolutional neural networks has recently been used as an approach to multimodal change detection. Existing approaches train the networks by exploiting supervised information of the change areas, which, however, is not always available. A main challenge in the unsupervised problem setting is to avoid that change pixels affect the learning of the translation function. We propose two new network architectures trained with loss functions weighted by priors that reduce the impact of change pixels on the learning objective. The change prior is derived in an unsupervised fashion from relational pixel information captured by domain-specific affinity matrices. Specifically, we use the vertex degrees associated with an absolute affinity difference matrix and demonstrate their utility in combination with cycle consistency and adversarial training. The proposed neural networks are compared with state-of-the-art algorithms. Experiments conducted on three real datasets show the effectiveness of our methodology.