AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

ChangeBridge: Spatiotemporal Image Generation with Multimodal Controls for Remote Senisng

Zhenghui Zhao, Chen Wu, Xiangyong Cao, Di Wang, Hongruixuan Chen, Datao Tang, Liangpei Zhang, Zhuo Zheng

CVPR 2026 · Image generation › Text-to-image diffusion · approach: Layout-to-image control

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Composed Bridge Initialization (start from composed pre-event state instead of noise) (ablation)ChangeBridge UNet (SD1.5) · WHU-CD instance layoutFID: 76.81 -> 56.24Improved
add Asynchronous Drift Diffusion on top of CompBridge (ablation)ChangeBridge UNet (SD1.5) · WHU-CD instance layoutFID: FID 56.24 -> 57.06; IS +0.18; IoU 71.87 -> 72.48Mixed
add Drift-Aware Denoising (ablation)ChangeBridge UNet (SD1.5) · WHU-CD instance layoutFID: 57.06 -> 45.47Improved
use DiT backbone instead of UNetChangeBridge · S2Looking instance layoutFID: 72.56 -> 56.42Improved
ChangeBridge versus Instruct-Imagen on coordinate textChangeBridge DiT · LEVIR-CCFID: 48.17 -> 31.45Improved
ChangeBridge versus Changen2 on semantic maskChangeBridge DiT · SECONDmIoU %: 73.20 -> 74.26Improved
add 1x synthetic data to real data for binary change detection (ablation)BiT · S2LookingIoU %: 46.94 -> 48.11Improved
add 2x synthetic data for binary change detection (ablation)BiT · WHU-CDIoU %: 72.39 -> 74.65Improved
increase synthetic data from 1x to 2x for semantic change detection (ablation)MambaSCD · SECONDmIoU %: 74.02 -> 73.57Hurt
increase synthetic data from 2x to 3x (ablation)BiT / RSICCformer · WHU-CD and LEVIR-CCIoU % / CIDEr-D: WHU 74.65 -> 73.40; CIDEr-D 145.09 -> 142.31Hurt
add synthetic data for change captioning (ablation)RSICCformer · LEVIR-CCCIDEr-D: 134.12 -> 145.09Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy