AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

SCFlow: Implicitly Learning Style and Content Disentanglement with Flow Models

Pingchuan Ma, Xiaopei Yang, Yusong Li, Ming Gui, Felix Krause, Johannes Schusterbauer, Björn Ommer

ICCV 2025 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Flow matching and flow sampling · 1 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
use SCFlow embeddings instead of CLIPCLIP embeddings · Style clustering, 51 stylesNMI: 0.4016 -> 0.8696Improved
use SCFlow embeddings instead of CLIPCLIP embeddings · Content clustering, 200 contentsNMI: 0.5368 -> 0.8356Improved
use SCFlow embeddings instead of CSDCSD style embeddings · Content clustering, 25 contentsNMI: 0.1888 -> 0.8374Improved
use SCFlow embeddings instead of CSDCSD style embeddings · Style clustering, 51 stylesFDR: 0.6328 -> 3.5184Improved
use SCFlow embeddings instead of CLIPCLIP embeddings · Style clustering, 51 stylesFDR: 0.2961 -> 3.5184Improved
use SCFlow content embeddings instead of CLIPCLIP embeddings · ImageNet-1k kNN (k=1)accuracy: 67.10 -> 66.25Hurt
use SCFlow style embeddings instead of CLIPCLIP embeddings · WikiArt retrievalRecall@1: 59.40 -> 65.34Improved
use SCFlow style embeddings instead of CSD-SCSD-S · WikiArt retrievalRecall@100: 95.58 -> 97.67Improved
train flow-matching merge model instead of contrastive loss on the same dataset (ablation)contrastive loss baseline · Style clustering, 51 stylesNMI: 0.2905 -> 0.8696Improved
train flow-matching merge model instead of InfoNCE on the same dataset (ablation)InfoNCE baseline · Style clustering, 51 stylesNMI: 0.5904 -> 0.8696Improved
train flow-matching merge model instead of InfoNCE on the same dataset (ablation)InfoNCE baseline · WikiArt style clusteringNMI: 0.4238 -> 0.4010Hurt
train flow-matching merge model instead of InfoNCE on the same dataset (ablation)InfoNCE baseline · ImageNet content clusteringFDR: 0.2056 -> 1.4264Improved

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