AI Paper Atlas

Research map / Image generation / Editing and style transfer

All-to-Key Attention for Arbitrary Style Transfer

Mingrui Zhu, Xiao He, Nannan Wang, Xiaoyu Wang, Xinbo Gao

ICCV 2023 · Image generation › Editing and style transfer · approach: Arbitrary neural style transfer · 63 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 all-to-all attention with all-to-key attention (ablation)StyA2K · image style transferContent Loss: 0.73 -> 0.55Improved
add distributed attention (ablation)StyA2K · image style transferStyle Loss: 1.48 -> 1.04Improved
add progressive attention (ablation)StyA2K · image style transferLPIPS: 0.59 -> 0.52Improved
add the coarse region-matching first step of progressive attention (ablation)StyA2K · image style transferContent Loss: content 0.65 -> 0.55; style 0.73 -> 1.04Mixed
use multiple attention heads instead of one (ablation)StyA2K · image style transferContent Loss: 0.65 -> 0.55Improved
increase lambda2 weight on attention-based feature matching loss (ablation)StyA2K · image style transferContent Loss vs Style Loss: content 0.72 -> 0.49; style 0.75 -> 1.38 (lambda2 0.5 -> 2)Mixed
StyA2K vs all-to-all methods in style lossStyA2K · image style transferStyle Loss: SANet 0.99 vs StyA2K 1.04Mixed
apply A2K to video frames without temporal trainingStyA2K · video style transfer (5 clips)optical flow error: 4.94 -> 4.24Improved
use sparse blocked attention instead of dense all-to-all attentionStyA2K · 512x512 inferenceinference time (sec/image): AdaAttN 0.0269 -> 0.0143Improved

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