AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

When Lines Meet Textures: Spatial-Frequency Aligned Diffusion Features for Cross-Sparsity Correspondence

Mingrui Zhu, Fengzhi Wang, Xin Wei, Jun Wang, Nannan Wang, Xinbo Gao

CVPR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Unimodal-to-multimodal representation transfer

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 LoFFA (low-frequency aggregation) to Unified CleanDIFT + Conv (ablation)SFA-DIFT · PSC6K+MS-PSC6K averagedPCK@5: 68.78 -> 69.83Improved
add AdaIN distribution alignment in LoFFA (ablation)SFA-DIFT · PSC6K+MS-PSC6K averagedPCK@5: 68.64 -> 69.83Improved
use LoFE instead of plain conv (ablation)SFA-DIFT · PSC6K+MS-PSC6K averagedPCK@5: 69.01 -> 69.83Improved
use DWT/IDWT instead of conv (ablation)SFA-DIFT · PSC6K+MS-PSC6K averagedPCK@5: 66.98 -> 69.83Improved
use two-level wavelet decomposition rather than single (ablation)SFA-DIFT · PSC6K+MS-PSC6K averagedPCK@5: 68.45 -> 69.83Improved
LoRA fine-tune CleanDIFT on sketches (Unified CleanDIFT), zero-shot (ablation)CleanDIFT · PSC6K+MS-PSC6K averagedPCK@5: 57.27 -> 58.69Improved
add DINOv2 features to Unified CleanDIFT, zero-shot (ablation)Unified CleanDIFT · PSC6K+MS-PSC6K averagedPCK@5: 58.69 -> 56.76Hurt
full SFA-DIFT vs CleanDIFT+DINO baselineCleanDIFT+DINO · PSC6KPCK@5: 70.25 -> 72.94Improved
full SFA-DIFT vs CleanDIFT+DINO on stylized dataCleanDIFT+DINO · MS-PSC6K AbstractPCK@5: 66.02 -> 70.43Improved
add LoFFA on top of Unified CleanDIFT for style robustness (ablation)Unified CleanDIFT · MS-PSC6K robustness ratioRR@1: 0.99 -> 0.87Hurt
fine-tune baselines on MS-PSC6KSelf-Sup / SketchFusion · MS-PSC6K AbstractPCK@5: Self-Sup 51.96 -> 45.23Hurt

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