AI Paper Atlas

Research map / Image generation / Editing and style transfer

Contrastive Diffusion Alignment: Learning Structured Latents for Controllable Generation

Ruchi Sandilya, Sumaira Perez, Charles Lynch, Lindsay Victoria, Benjamin Zebley, Derrick Matthew Buchanan, Mahendra T. Bhati, Nolan Williams, Timothy Spellman, FAITH M. GUNNING, Conor Liston, Logan Grosenick

ICML 2026 · Image generation › Editing and style transfer · approach: Inversion-based diffusion editing

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
interpolate in the contrastive embedding C instead of raw diffusion latent Z (TEX-1) (ablation)TEX-1 · Fluid interpolationPSNR: 29.28 -> 32.98Improved
use C-space instead of Z-space for LSTM RMSE (ablation)LSTM · Fluid interpolationRMSE: 4.04 -> 1.07Improved
use C-space instead of Z-space for TEX-2 (ablation)TEX-2 · Fluid interpolationPSNR: 35.29 -> 35.70Improved
linear and spherical interpolation (Lerp/Slerp) in either space (ablation)Lerp/Slerp · Fluid interpolationPSNR: 28.16 -> 28.27No clear effect
second-order Taylor term (TEX-2) instead of first-order (TEX-1) (ablation)TEX extrapolation · Ca2+ interpolation (C-space)PSNR: 36.13 -> 38.58Improved
n-step-ahead prediction with spline vs LSTM vs TEX-2 in C-space (ablation)ConDA · Ca2+ n-step-aheadPSNR: 35.11 (LSTM) / 34.92 (TEX-2) / 35.00 (Spline)No clear effect
train SVM on ConDA embedding rather than raw diffusion latent (E-field responder) (ablation)SVM-RBF · Responder vs non-responderF1: 0.30 -> 0.63Improved
train SVM on ConDA embedding rather than raw diffusion latent (flow) (ablation)SVM-RBF · Steady vs unsteady flowaccuracy: 0.68 -> 0.87Improved
use contrastive ConDA embedding instead of PCA (ablation)ConDA · Fluid held-out segment predictionPSNR: 31.82 -> 33.90Improved
use contrastive ConDA embedding instead of beta-VAE (ablation)ConDA · Fluid held-out segment predictionRMSE: 7.47 -> 7.60Mixed
use contrastive ConDA embedding instead of beta-VAE (ablation)ConDA · DISFA held-out segment predictionPSNR: 35.76 -> 36.26Improved
kNN decoder lifting instead of tuned MLP decoder (ablation)ConDA · multiple datasetsreconstructionImproved

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