AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

UFC-BERT: Unifying Multi-Modal Controls for Conditional Image Synthesis

Zhu Zhang, Jianxin Ma, Chang Zhou, Rui Men, Zhikang Li, Ming Ding, Jie Tang, Jingren Zhou, Hongxia Yang

NeurIPS 2021 · Image generation › Autoregressive and latent generation · approach: Generative latent space tooling · 23 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 relevance estimator (ablation)PNAG · M2C-FashionFID: 12.17 -> 11.53Improved
add fidelity estimator (ablation)PNAG · M2C-FashionFID: 13.14 -> 11.53Improved
add relevance estimator (ablation)PNAG · Multi-Modal CelebA-HQFID: 68.90 -> 66.72Improved
add fidelity estimator (ablation)PNAG · Multi-Modal CelebA-HQFID: 70.32 -> 66.72Improved
use PNAG with both estimators instead of Mask-Predict (ablation)UFC-BERT · M2C-FashionFID: 14.77 -> 11.53Improved
use PNAG instead of Mask-Predict (ablation)UFC-BERT · Multi-Modal CelebA-HQFID: 72.04 -> 66.72Improved
increase parallel decoding number B from 1 to 5 (ablation)PNAG · M2C-FashionFID: 12.72 -> 11.53Improved
increase B from 5 to 10 (ablation)PNAG · Multi-Modal CelebA-HQFID: 66.72 -> 65.30Improved
replace autoregressive VQGAN with non-autoregressive UFC-BERTtwo-stage synthesis · M2C-FashionFID: 12.48 -> 11.53Improved
replace autoregressive VQGAN with non-autoregressive UFC-BERTtwo-stage synthesis · Multi-Modal CelebA-HQFID: 52.63 -> 66.72Hurt
replace autoregressive VQGAN with non-autoregressive UFC-BERTtwo-stage synthesis · M2C-Fashioninference time: 8.73 sec/sample -> 0.81 sec/sampleImproved
use UFC-BERT instead of TediGANtext-to-image · Multi-Modal CelebA-HQFID: 106.37 -> 66.72Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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