ChangeFlow - Latent Rectified Flow for Change Detection in Remote Sensing
Abstract
Remote sensing change detection (RSCD) localises changes between two images of the same geographic region. Most state-of-the-art methods are trained with a per-pixel discriminative objective that classifies each spatial location independently. In this scenario, the predicted changed region is not modelled as a coherent whole, so predictions tend to be spatially fragmented. Generative modelling offers a principled solution: by learning a distribution over plausible change masks, it treats the mask as a single object and encourages global consistency. Yet existing generative RSCD methods lag behind strong discriminative baselines, held back by costly pixel-space generation and overly complex conditioning. We introduce ChangeFlow, which reformulates change detection as the generative synthesis of change masks in a compact latent space via rectified flow, guided by a structured yet lightweight bi-temporal conditioning signal. Changeflow yields spatially coherent predictions without sacrificing efficiency: across four binary benchmarks, SYSU, LEVIR, CLCD, and OSCD, ChangeFlow achieves an average F1 of 80.4%, a 1.3-point gain over the previous best with better efficiency. It also extends to semantic change detection, setting a new state-of-the-art 65.9 Fscd on SECOND.
Mask Generation Examples
Contributions
- As our main contribution, we reformulate RSCD as latent-space change mask generation and propose a rectified flow framework that jointly models change masks and produces globally coherent predictions.
- We show that auxiliary predictors and complex layers are unnecessary for conditioning and propose a simpler strategy based on feature difference.
- We demonstrate the generality of the formulation by extending ChangeFlow to semantic change detection.
ChangeFlow
Given a pair of images, we first extract features using a Shared Weight Encoder, and we condition the Diffusion Transformer (DiT) rectified flow model on the absolute difference of the extracted features. Guided by this conditioning, the model then iteratively generates a latent representation of the corresponding change mask, which is ultimately decoded by the Variational Autoencoder (VAE) into a binary change mask.
Results
False positives are marked in red and false negatives in blue.
Semantic CD Results
Results on the SECOND datasets using the semantic version of ChangeFlow.
Coherence Analysis
Coherence measured as hole count error (lower is better) and boundary F1 (BF1, higher is better) averaged over 4 datasets. ChangeFlow yields low structural error, indicating the fewest spurious holes and high border precision.
Number of steps and sampling based inference
Impact of number of sampling steps (at fixed rate of 5 repetitions) and inference repetitions (at fixed 5 sampling steps). Change detection performance is reported on the left y-axis, measured as average F1 across 4 datasets, while inference speed is reported on the right y-axis as frames per second. Sampling-based ensembling (multiple repetitions), enabled by generative formulation, provides a controllable speed–accuracy trade-off at inference time.
BibTeX
@article{rolih2026changeflow,
title={ChangeFlow - Latent Rectified Flow for Change Detection in Remote Sensing},
author={Rolih, Blaž and Fučka, Matic and Wolf, Filip and Čehovin Zajc, Luka},
journal={arXiv},
year={2026},
eprint={2605.15375},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2605.15375},
}