Research map / Video / Video generation
Long-Term Temporally Consistent Unpaired Video Translation From Simulated Surgical 3D Data
Dominik Rivoir, Micha Pfeiffer, Reuben Docea, Fiona Kolbinger, Carina Riediger, Jürgen Weitz, Stefanie Speidel
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 neural textures (without view-consistency loss) (ablation) | Ours w/o tex,vc · Simulated surgical scenes | ORB-1 match accuracy: 81.7% -> 88.3% | Improved |
| add neural textures (without view-consistency loss) (ablation) | Ours w/o tex,vc · Simulated surgical scenes | ORB-10 match accuracy: 29.5% -> 44.5% | Improved |
| add neural textures (without view-consistency loss) (ablation) | Ours w/o tex,vc · Simulated surgical scenes | optical-flow error (OF): 8.43 -> 8.35 | Improved |
| add neural textures (without view-consistency loss) (ablation) | Ours w/o tex,vc · Simulated surgical scenes | FID: 27.3 -> 27.0 | No clear effect |
| add neural textures (without view-consistency loss) (ablation) | Ours w/o tex,vc · Simulated surgical scenes | liver segmentation Dice (%): 56.8 -> 55.2 | Hurt |
| add lighting-invariant view-consistency loss (ablation) | Ours w/o vc · Simulated surgical scenes | optical-flow error (OF): 8.35 -> 7.62 | Improved |
| add lighting-invariant view-consistency loss (ablation) | Ours w/o vc · Simulated surgical scenes | ORB-10 match accuracy: 44.5% -> 49.6% | Improved |
| add lighting-invariant view-consistency loss (ablation) | Ours w/o vc · Simulated surgical scenes | ORB-1 match accuracy: 88.3% -> 91.8% | Improved |
| remove style and noise injection from the generator (SSIM-MUNIT to Ours w/o tex,vc) (ablation) | SSIM-MUNIT · Simulated surgical scenes | ORB-1 match accuracy: 60.5% -> 81.7% | Improved |
| add MS-SSIM loss for label preservation (ReCycle to SSIM-ReCycle) | ReCycle · Simulated surgical scenes | liver segmentation Dice (%): 40.7 -> 50.9 | Improved |
| add MS-SSIM loss for label preservation (ReCycle to SSIM-ReCycle) | ReCycle · Simulated surgical scenes | FID: 61.5 -> 80.6 | Hurt |
| revisit earlier views instead of consecutive frames | OF-UNIT · Simulated surgical scenes | ORB-1 match accuracy: 93.5% -> 69.9% | Hurt |
Closest papers by meaning
- TC-Light: Temporally Coherent Generative Rendering for Realistic World Transfer (NeurIPS 2025 · similarity 0.85)
- 4Real: Towards Photorealistic 4D Scene Generation via Video Diffusion Models (NeurIPS 2024 · similarity 0.85)
- SimVS: Simulating World Inconsistencies for Robust View Synthesis (CVPR 2025 · similarity 0.84)
- MultiDiff: Consistent Novel View Synthesis from a Single Image (CVPR 2024 · similarity 0.84)
- Consistent4D: Consistent 360° Dynamic Object Generation from Monocular Video (ICLR 2024 · similarity 0.84)
- SurgVista: Long-Horizon Surgical World Modeling with Plausible Instrument-Tissue Dynamics (NeurIPS 2026 · similarity 0.84)
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.
