AI Paper Atlas

Research map / Image generation / Image restoration

SDiD:Shared diffusion prior for efficient distributed stereo image compression

Yichong Xia, Yimin Zhou, Zongyu Li, Shiyu Qin, Mingyao Hong, Bin Chen, Haoqian Wang

ICML 2026 · Image generation › Image restoration · approach: Learned image compression

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 diffusion prior alignment (ablation)SDiD · InStereo2KBD-rate LPIPS (main view): 113.52Improved
add diffusion prior alignment (ablation)SDiD · InStereo2KBD-rate RMSE (aux view): 152.15Improved
add adaptive noise estimator (ablation)SDiD · InStereo2KBD-rate LPIPS (main view): 21.53Improved
add adaptive noise estimator (ablation)SDiD · InStereo2KBD-rate LPIPS (aux view): 25.20Improved
add diffusion prior transfer (ablation)SDiD · InStereo2KBD-rate LPIPS (aux view): 232.71Improved
add look-back attention over standard cross-attention (ablation)SDiD · InStereo2KBD-rate LPIPS (aux view): 25.50Improved
add perceptual loss Lper (ablation)SDiD · InStereo2KBD-rate LPIPS: 80.22Improved
add perceptual loss Lper (ablation)SDiD · InStereo2KBD-rate RMSE: 30.11Improved
add DINOv2 feature loss Ldino (ablation)SDiD · InStereo2KBD-rate LPIPS: 15.31Improved
add DINOv2 feature loss Ldino (ablation)SDiD · InStereo2KBD-rate RMSE: 35.26Improved
add adversarial loss Ladv (ablation)SDiD · InStereo2KBD-rate LPIPS: 27.18Improved
add adversarial loss Ladv (ablation)SDiD · InStereo2KBD-rate RMSE: 15.45Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 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