Research map / Image generation / GANs
Conditional Visual Autoregressive Modeling for Pathological Image Restoration
Ziyi Liu, Zhe Xu, Jiabo Ma, Wenqiang Li, Ruixuan Wang, Bo Du, Hao Chen
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add conditional tokens to VAR baseline (ablation) | CVARPath · MIDOG2022 SR x8 | FID: 124.28 -> 96.46 | Improved |
| add conditional tokens to VAR baseline (ablation) | CVARPath · MIDOG2022 SR x8 | PSNR: 17.15 -> 21.67 | Improved |
| add attention mask on top of conditional tokens (ablation) | CVARPath · MIDOG2022 SR x8 | FID: 96.46 -> 52.72 | Improved |
| add attention mask on top of conditional tokens (ablation) | CVARPath · MIDOG2022 SR x8 | PSNR: 21.67 -> 23.48 | Improved |
| add contrastive pretraining of the VQVAE encoder (ablation) | CVARPath · MIDOG2022 SR x8 | FID: 52.72 -> 41.10 | Improved |
| add contrastive pretraining of the VQVAE encoder (ablation) | CVARPath · MIDOG2022 SR x8 | PSNR: 23.48 -> 24.22 | Improved |
| use only 8x8 degraded sparse tokens as condition (ablation) | CVARPath · MIDOG2022 (MAE vs token scale) | MAE | Improved |
| CVARPath vs SwinIR on denoising density 0.3 | CVARPath · MIDOG2022 denoising density 0.3 | PSNR: 30.299 -> 28.377 | Hurt |
| CVARPath vs best baseline on SR x8 FID | CVARPath · MIDOG2022 SR x8 | FID: 111.173 -> 41.100 | Improved |
| CVARPath vs best baseline on SR x8 FID | CVARPath · TIGER2021 SR x8 | FID: 96.294 -> 78.119 | Improved |
| CVARPath vs CycleGAN on deblurring kernel 7 PSNR/SSIM | CVARPath · MIDOG2022 deblurring kernel 7 | SSIM: 0.879 -> 0.871 | Hurt |
| CVARPath vs SwinIR on denoising density 0.5 | CVARPath · MIDOG2022 denoising density 0.5 | FID: 90.466 -> 42.514 | Improved |
Closest papers by meaning
- PathVQ: Reforming Computational Pathology Foundation Model for Whole Slide Image Analysis via Vector Quantization (NeurIPS 2025 · similarity 0.88)
- FOCUS: Knowledge-enhanced Adaptive Visual Compression for Few-shot Whole Slide Image Classification (CVPR 2025 · similarity 0.86)
- Multi-Resolution Pathology-Language Pre-training Model with Text-Guided Visual Representation (CVPR 2025 · similarity 0.86)
- PathGen-1.6M: 1.6 Million Pathology Image-text Pairs Generation through Multi-agent Collaboration (ICLR 2025 · similarity 0.85)
- PathView-Bench: Can Multimodal Large Language Models Achieve Fine-grained Multiscale Understanding of Pathology Images? (NeurIPS 2026 · similarity 0.85)
- Autoregressive Sequence Modeling for 3D Medical Image Representation (AAAI 2025 · similarity 0.85)
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.
