AI Paper Atlas

Research map / Detection and segmentation / CNN features

Virtual Immunohistochemistry Staining with Dual-Aligned Multi-Task Feature Guidance

Shigeng Xie, Hongming Xu, Guiyang Jiang, Tuomo Rossi, Tommi Kärkkäinen, Fengyu Cong

CVPR 2026 · Detection and segmentation › CNN features · approach: Domain-specific segmentation and detection

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 unaligned multi-task feature guidance directly to the VIS generator (ablation)Virtual IHC staining generator · MIST-HER2FID: 46.23 -> 74.50Hurt
add spatial alignment to feature guidance (no task gap alignment) (ablation)Feature-guided VIS · MIST-HER2FID: 74.50 -> 50.99Improved
add task gap alignment model to feature guidance (no spatial alignment) (ablation)Feature-guided VIS · MIST-HER2FID: 74.50 -> 44.60Improved
add both spatial and task gap alignment (full method) (ablation)Feature-guided VIS · MIST-HER2FID: 46.23 -> 40.34Improved
add structure-enhanced learning (SEL) (ablation)Spatial alignment · MIST-HER2FID: 44.59 -> 40.34Improved
use active-passive matching instead of global matching (ablation)Spatial alignment · MIST-HER2FID: 44.97 -> 40.34Improved
add H&E auxiliary task features alongside IHC features (ablation)Multi-task guidance · MIST-HER2FID: 43.12 -> 40.34Improved
add reconstruction task features alongside classification features (ablation)Multi-task guidance · MIST-HER2FID: 42.31 -> 40.34Improved
use proposed spatial alignment instead of optimal-transport matching (ablation)Spatial alignment · MIST-HER2FID: 43.72 -> 40.34Improved
proposed method vs best baseline on SSIMMIST datasets · MIST-HER2SSIM: 0.1905 vs CUT 0.1708 / StegoGAN 0.2101No clear effect
proposed method vs SIM-GAN and PSPStainBCI-HER2 · BCI-HER2FID: 47.86 -> 45.57Improved

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