AI Paper Atlas

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

Zero-Shot Contrastive Loss for Text-Guided Diffusion Image Style Transfer

Serin Yang, Hyunmin Hwang, Jong Chul Ye

ICCV 2023 · Image generation › Text-to-image diffusion · approach: Text-driven synthesis and semantic guidance · 99 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 ZeCon patch-wise contrastive content loss to MSE and VGG content losses (ablation)ZeCon · user study (content)user study content score: 3.29 -> 4.81Improved
use ZeCon loss alone instead of MSE and VGG combination (ablation)ZeCon · user study (content)user study content score: 3.29 -> 4.29Improved
add directional CLIP loss to global CLIP loss (ablation)ZeCon style guidance · qualitative stylesstylization strengthImproved
use patch-based CLIP guidance instead of whole-image guidance (ablation)ZeCon style guidance · qualitative stylesstylization coverageImproved
add augmentation and use N=32 patches (N<32 insufficient) (ablation)patch CLIP loss · Golden stylestylization: inference time 24 seconds with N=32Improved
increase t0 for the forward diffusion (ablation)ZeCon · style transferCLIP score vs content preservationMixed
use ZeCon instead of DiffusionCLIP for inferenceDiffusionCLIP · computational costinference time (sec): 96 -> 38Improved
apply method to unseen-domain imagesDiffusionCLIP · user study content scorecontent score: DiffusionCLIP 3.71 -> 3.29 (photo -> unseen); ZeCon 4.76 -> 4.62Improved
use ZeCon instead of StyleCLIP on face identity lossStyleCLIP · Face IDFace ID (lower better): 0.3750 -> 0.3881Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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