AI Paper Atlas

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

Stochastic Interpolants for Revealing Stylistic Flows across the History of Art

Pingchuan Ma, Ming Gui, Johannes Schusterbauer, Xiaopei Yang, Olga Grebenkova, Vincent Tao Hu, Björn Ommer

ICCV 2025 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Flow matching and flow sampling · 0 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
fine-tune Stable Diffusion 1.5 on the art dataset with year promptsSD 1.5 · Feature synthesis, 16th centuryFID: 131.1 -> 111.7Improved
fine-tune Stable Diffusion 1.5 on the art dataset with year promptsSD 1.5 · Feature synthesis, 20th centuryFID: 180.4 -> 150.7Improved
fine-tune SDEdit on the art datasetSDEdit · Time matching, Rococo artworksRecall@1: 6.3 -> 37.9Improved
fine-tune SDEdit on the art datasetSDEdit · Time matching, Realism artworksRecall@1: 3.7 -> 27.5Improved
use latent style-space stochastic interpolant bridge instead of fine-tuned SDEditSDEdit (fine-tuned) · Time matching, Rococo artworksRecall@1: 37.9 -> 72.5Improved
use latent style-space bridge instead of fine-tuned SDEditSDEdit (fine-tuned) · Time matching, Impressionism artworkscompactness: 0.272 -> 0.024Improved
use latent bridge instead of SDEditSDEdit · Time matching, Impressionism artworkstriplet consistency (%): 34.5 -> 75.8Improved
use latent bridge instead of RF-inversionRF-inversion · Time matching, Rococo artworkstriplet consistency (%): 83.7 -> 84.8Mixed
use the model with year conditioning instead of text-conditioned SDXLSDXL · Feature synthesis, 19th centuryFID: 114.8 -> 72.0Improved
use the model instead of FLUX.1-schnellFLUX.1-schnell · Feature synthesis, 20th centuryFID: 126.1 -> 78.9Improved
set integration threshold t_th too low (ablation)stochastic interpolant bridge · qualitativecontent preservationHurt
raise integration threshold t_th toward 1 (ablation)stochastic interpolant bridge · qualitativesemantic identityImproved

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