AI Paper Atlas

Research map / Image generation / Image restoration

RevealLayer: Disentangling Hidden and Visible Layers via Occlusion-Aware Image Decomposition

Binhao Wang, Shihao Zhao, Bo Cheng, Qiuyu Ji, Yuhang Ma, Liebucha Wu, Shanyuan Liu, Dawei Leng, Yuhui Yin

ICML 2026 · Image generation › Image restoration · approach: Dehazing and deraining · 0 citations (OpenAlex)

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 Region-Aware Attention (RAA) (ablation)RevealLayer baseline · RevealLayer-100K backgroundPSNR: 22.08 -> 23.80Improved
add RAA (ablation)RevealLayer baseline · RevealLayer-100K foregroundFID: 20.86 -> 19.66Improved
add Occlusion-Guided Adapter (OGA) to RAA (ablation)RevealLayer · RevealLayer-100K backgroundPSNR: 23.80 -> 24.97Improved
add orthogonality loss (ablation)RAA + OGA · RevealLayer-100K backgroundFID: 57.51 -> 53.63Improved
add orthogonality loss (ablation)RAA + OGA · RevealLayer-100K foregroundSoftIoU: 0.9271 -> 0.9418Improved
add alpha loss (ablation)RAA + OGA · RevealLayer-100K foregroundSoftIoU: 0.9271 -> 0.9420Improved
add alpha loss to orthogonality-loss variant (ablation)RAA + OGA + OrthLoss · RevealLayer-100K backgroundPSNR: 24.96 -> 25.53Improved
share the same noise across foreground layers at inference (ablation)RevealLayer · RevealLayerBench foregroundPSNR: 32.13 -> 32.31Improved
5%-10% offset in input bounding box (ablation)RevealLayer · RevealLayerBench foregroundSoftIoU: 0.9432 -> 0.8569Hurt
enlarge input box by 10%-20% (ablation)RevealLayer · RevealLayerBench backgroundPSNR: 25.53 -> 25.30Hurt
fine-tune VAE decoder on RevealLayer-100K (XVAE) vs original ART VAE (ablation)RevealLayer · natural-scene background reconstructionPSNR: 46.68 -> 48.74Improved
fine-tune RevealLayer 4k steps on PrismLayers vs CLDRevealLayer · PrismLayersPro validationPSNR / FID: PSNR 27.646 -> 28.360; FID 19.413 -> 22.716Mixed

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