AI Paper Atlas

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

Discrete Noise Inversion for Next-scale Autoregressive Text-based Image Editing

Minh Quan Dao, Xiaoxiao He, Ligong Han, Ngan Nguyen, Amin Nobari, Han Zhang, Faez Ahmed, Viet Anh Nguyen, Dimitris N. Metaxas

ECCV 2026 · Image generation › Text-to-image diffusion · approach: Text-driven synthesis and semantic guidance · 7 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
replace Regeneration with VARIN noise inversionHART · PIE-Bench editingStructure distance x10^3: 25.56 -> 11.46Improved
replace Regeneration with VARIN noise inversionHART · PIE-Bench editingPSNR: 20.45 -> 26.54Improved
replace Regeneration with VARIN noise inversionHART · PIE-Bench editingCLIP similarity (edited): 21.13 -> 21.49Hurt
use location-aware inversion with VARIN in HART instead of DICE on PaellaVARIN · PIE-Bench editingCLIP similarity (whole): 23.79 -> 25.05Improved
VARIN with HART instead of DICE with PaellaVARIN · PIE-Bench editingPSNR: 27.29 -> 26.54Hurt
VARIN instead of EditARVARIN · PIE-Bench editingPSNR: 21.32 -> 26.54Improved
increase tau from 14 to 20 in LAI (ablation)VARIN · PIE-Bench tau ablationPSNR: 23.97 -> 27.30Improved
increase tau from 14 to 20 in LAI (ablation)VARIN · PIE-Bench tau ablationCLIP similarity (whole): 25.73 -> 24.79Hurt
increase Regeneration start scale s from 6 to 8 (ablation)Regeneration · PIE-Bench Regeneration start scalePSNR: 19.21 -> 20.45Improved
increase Regeneration start scale s from 6 to 8 (ablation)Regeneration · PIE-Bench Regeneration start scaleCLIP similarity (whole): 26.83 -> 24.65Hurt
use VARIN instead of DDPM-Inversion (user study)VARIN · User studyprompt agreement %: 52.35% -> 64.71%Improved

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