Research map / Video / Video generation
DiffuPrompt: Adapting Video Foundation Models to 3D Medical Volumes via Latent Trajectory Priors
Guowei Dai, Duwei Dai, yulong ji, Chen Hu, Yi Zhang
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 static slice-wise latent priors (ablation) | Vanilla SAM2 (FIFO memory) · BTCV (13 organs) | DSC: 58.4 -> 63.2 | Improved |
| add LLM anatomical semantics to the prior (ablation) | DiuPrompt v1 · BTCV (13 organs) | DSC: 63.2 -> 68.5 | Improved |
| replace static prior with continuous Neural ODE trajectory (ablation) | DiuPrompt v2 · BTCV (13 organs) | DSC / HD95: DSC 68.5 -> 74.6; HD95 12.80 -> 8.90 | Improved |
| replace FIFO memory with dual-stream memory (simple addition) (ablation) | DiuPrompt v3 · BTCV (13 organs) | DSC: 74.6 -> 78.9 | Improved |
| add learned spatial confidence gate instead of simple addition (ablation) | DiuPrompt v4 · BTCV (13 organs) | DSC / HD95: DSC 78.9 -> 82.7; HD95 6.20 -> 4.46 | Improved |
| DiuPrompt vs strongest promptable baseline AoP-SAM | SAM2 promptable segmentation · MSD (10-task macro-avg) | HD95: 5.75 -> 4.92 mm | Improved |
| DiuPrompt vs fully supervised nnU-Net | BTCV (13 organs) · BTCV (13 organs) | DSC: 86.2 -> 82.7 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SAM2Long — SAM2Long: Enhancing SAM 2 for Long Video Segmentation with a Training-Free Memory Tree (ICCV 2025) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Learning General-purpose Biomedical Volume Representations using Randomized Synthesis (ICLR 2025 · similarity 0.87)
- UniPro: Unified Multi-Mode Medical Image Segmentation from 2D Images to 3D Volumes via Propagation (NeurIPS 2026 · similarity 0.87)
- LiFT: Lifted Inter-slice Feature Trajectories for 3D Image Generation from 2D Generators (NeurIPS 2026 · similarity 0.86)
- VISTA3D: A Unified Segmentation Foundation Model For 3D Medical Imaging (CVPR 2025 · similarity 0.86)
- Introducing 3D Representation for Dense Volume-to-Volume Translation via Score Fusion (ICML 2025 · similarity 0.85)
- Knowledge Transfer Scaling Laws for 3D Medical Imaging (NeurIPS 2026 · 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.
