AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codes

Sam Bond-Taylor, Peter Hessey, Hiroshi Sasaki, Toby P. Breckon, Chris G. Willcocks

ECCV 2022 · Image generation › Autoregressive and latent generation · approach: High-resolution diffusion synthesis · 61 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 autoregressive prior with absorbing diffusion (ablation)VQ image prior · LSUN ChurchesFID: 5.93 -> 5.58Improved
replace autoregressive prior with absorbing diffusion (ablation)VQ image prior · FFHQFID: 8.15 -> 7.12Improved
replace autoregressive prior with absorbing diffusion (ablation)VQ image prior · LSUN ChurchesNLL: 6.24 -> 6.01Improved
replace autoregressive prior with absorbing diffusion (ablation)VQ image prior · FFHQNLL: 6.18 -> 5.96Improved
increase sampling steps from 50 to 256 (ablation)Unleashing Transformers · LSUN ChurchesFID: 6.86 -> 5.58Improved
increase sampling steps from 50 to 256 (ablation)Unleashing Transformers · FFHQFID: 9.60 -> 7.12Improved
use sampling steps 200 instead of 256 (ablation)Unleashing Transformers · LSUN BedroomFID: 5.42 -> 5.32No clear effect
add DiffAug only to VQ model training (ablation)VQ image model · LSUN Churchesreconstruction FID: 5.25 -> 5.16Mixed
add DiffAug only to VQ model training (ablation)VQ image model · FFHQreconstruction FID: 3.37 -> 6.57Hurt
add adaptive weight limit lambda_max=1 only (ablation)VQ image model · LSUN Churchesreconstruction FID: 5.25 -> 8.67Hurt
add both DiffAug and adaptive weight limiting (ablation)VQ image model · LSUN Churchesreconstruction FID: 5.25 -> 2.70Improved
add both DiffAug and adaptive weight limiting (ablation)VQ image model · FFHQreconstruction FID: 3.37 -> 3.12Improved

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