Research map / Probabilistic and scientific modeling / Dynamical systems
Progression as Latent Drift: Generative Forecasting of Slow-Evolving Pathologies
Yuxiang Feng, Juncheng Wang, Chao Xu, Wenlong Hou, Huihan Wang, Yijie Qian, Yang Liu, Baigui Sun, Yong Liu, Shujun Wang
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 |
|---|---|---|---|
| predict latent drift instead of pixel-level future (ablation) | Latent Drift · ADNI+AIBL forecasting | FID (i3d): 44.38 -> 13.92 | Improved |
| predict latent drift instead of pixel-level future (ablation) | Latent Drift · ADNI+AIBL forecasting | Diff-SSIM: 0.7625 -> 0.8205 | Improved |
| predict latent drift instead of pixel-level future (ablation) | Latent Drift · ADNI+AIBL forecasting | Accuracy: 77.19 -> 88.33 | Improved |
| use FSQ instead of vector quantization (ablation) | Latent Drift · ADNI+AIBL forecasting | Accuracy: 86.67 -> 88.24 | Improved |
| use FSQ instead of residual VQ (ablation) | Latent Drift · ADNI+AIBL forecasting | Accuracy: 84.56 -> 88.24 | Improved |
| use FSQ instead of lookup-free quantization (ablation) | Latent Drift · ADNI+AIBL forecasting | F1: 84.02 -> 87.41 | Improved |
| use FSQ instead of BSQ (ablation) | Latent Drift · ADNI+AIBL forecasting | F1: 82.75 -> 87.41 | Improved |
| use FSQ instead of VQ or LFQ for structural similarity (ablation) | Latent Drift · ADNI+AIBL forecasting | Diff-SSIM: 0.8249 -> 0.8224 | Hurt |
| use FSQ grid [8,8,8,5,5,5] instead of [8,8,8,6,5] (ablation) | Latent Drift · ADNI+AIBL forecasting | rFID: 27.07 -> 12.78 | Improved |
| use FSQ grid [8,8,8,5,5,5] instead of coarse [8,6,5] (ablation) | Latent Drift · ADNI+AIBL forecasting | Accuracy: 86.84 -> 88.24 | Improved |
| use autoregressive generator instead of context-conditioning (ablation) | Latent Drift · ADNI+AIBL forecasting | Accuracy: 83.77 -> 88.33 | Improved |
| use autoregressive generator instead of masked-conditioning (ablation) | Latent Drift · ADNI+AIBL forecasting | rFID: 23.17 -> 13.92 | Improved |
Closest papers by meaning
- Learning Patient-Specific Disease Dynamics With Latent Flow Matching For Longitudinal Imaging Generation (ICLR 2026 · similarity 0.89)
- Latent Drifting in Diffusion Models for Counterfactual Medical Image Synthesis (CVPR 2025 · similarity 0.89)
- Unscrambling disease progression at scale: fast inference of event permutations with optimal transport (NeurIPS 2024 · similarity 0.86)
- Adaptive Shrinkage Estimation for Personalized Deep Kernel Regression in Modeling Brain Trajectories (ICLR 2025 · similarity 0.86)
- Generative Modeling via Drifting (NeurIPS 2026 · similarity 0.86)
- A scalable generative model for dynamical system reconstruction from neuroimaging data (NeurIPS 2024 · 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.
