AI Paper Atlas

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

TextStyleBrush: Transfer of Text Aesthetics From a Single Example

Praveen Krishnan, Rama Kovvuri, Guan Pang, Boris Vassilev, Tal Hassner

TPAMI 2023 · Image generation › Text-to-image diffusion · approach: Personalized and region-controlled generation · 82 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 reconstruction loss l_rec (ablation)l_D + l_R model · Synth-Paired ablationFID: 165.41 -> 86.79Improved
add cyclic reconstruction loss l_cyc (ablation)l_D + l_R + l_rec model · Synth-Paired ablationSSIM: 0.4164 -> 0.4434Improved
add cyclic reconstruction loss l_cyc (ablation)l_D + l_R + l_rec model · Synth-Paired ablationFID: 86.79 -> 99.45Hurt
add typeface perceptual loss l_type (ablation)l_D + l_R + l_rec + l_cyc model · Synth-Paired ablationFID: 99.45 -> 79.49Improved
add typeface perceptual loss l_type (ablation)l_D + l_R + l_rec + l_cyc model · Synth-Paired ablationSSIM: 0.4434 -> 0.4174Hurt
use multi-scale style representation instead of single-scale (ablation)TSB full losses · Synth-Paired ablationFID: 87.68 -> 79.49Improved
add foreground mask learning (ablation)TSB full losses · Synth-Paired ablationFID: 90.67 -> 79.49Improved
add foreground mask learning (ablation)TSB full losses · Synth-Paired ablationSSIM: 0.4416 -> 0.4174Hurt
use TSB instead of SRNetscene text generation · IC15 word imagesrecognition accuracy: 23.4 -> 97.6Improved
use TSB instead of Davis et al.handwriting generation · IAM wordsFID: 104.95 -> 44.68Improved
use TSB instead of SRNet in a user realism testscene text generation · User study scene textpercent judged real: 28.64 -> 58.88Improved
use TSB instead of Davis et al. in a user realism testhandwriting generation · User study handwritingpercent judged real: 37.45 -> 65.76Improved

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