AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Draft-and-Revise: Effective Image Generation with Contextual RQ-Transformer

Doyup Lee, Chiheon Kim, Saehoon Kim, Minsu Cho, Wook-Shin Han

NeurIPS 2022 · Image generation › Autoregressive and latent generation · approach: Next-scale and token autoregressive images · 6 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 classifier-free guidance only in the revise phase (ablation)Contextual RQ-Transformer 821M · ImageNetFID: 5.78 -> 3.45Improved
apply guidance in draft phase too (ablation)Contextual RQ-Transformer 821M · ImageNetFID: 3.45 -> 8.90Hurt
apply guidance in draft phase too (ablation)Contextual RQ-Transformer 821M · ImageNetRecall: 0.52 -> 0.33Hurt
increase T_draft from 8 to 64 (ablation)Contextual RQ-Transformer · ImageNetFID: 5.41 -> 3.45Improved
increase T_draft from 8 to 64 (ablation)Contextual RQ-Transformer · ImageNets/sample: 0.03 -> 0.10Hurt
increase number of parameters from 821M to 1.4BContextual RQ-Transformer · ImageNetFID: 3.45 -> 3.41No clear effect
increase parameters from 371M to 821MContextual RQ-Transformer · ImageNetFID: 5.45 -> 3.45Improved
increase T_revise (ablation)Contextual RQ-Transformer 821M · ImageNetFIDHurt
revise phase after draft phase (ablation)Contextual RQ-Transformer · ImageNetFID: 15.32 -> 3.45Improved
Top-C confidence-based infilling instead of random in draft (T_draft=64) (ablation)Contextual RQ-Transformer 821M · ImageNetFID: 3.45 -> 27.88Hurt
Top-C-50% instead of random in draft (T_draft=64) (ablation)Contextual RQ-Transformer 821M · ImageNetFID: 3.45 -> 6.84Hurt
Top-C-50% instead of random (T_draft=8) (ablation)Contextual RQ-Transformer 821M · ImageNetFID: 6.28 -> 6.13Improved

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