Research map / 3D vision / Depth and stereo
3D Video Stabilization With Depth Estimation by CNN-Based Optimization
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 geometric loss and optical-flow loss to the reconstruction objective (ablation) | Deep3D Stabilizer · NUS | stability: 16.95 -> 18.29 | Helped |
| Add geometric loss to the objective that already includes photometric loss (ablation) | Deep3D Stabilizer · NUS | stability: 17.35 -> 18.29 | Helped |
| Add geometric loss and optical-flow loss to the reconstruction objective (ablation) | Deep3D Stabilizer · NUS | global distortion: 0.87 -> 0.92 | Helped |
| Add geometric loss to the objective that already includes photometric loss (ablation) | Deep3D Stabilizer · NUS | global distortion: 0.89 -> 0.92 | Helped |
| Add geometric loss and optical-flow loss to the reconstruction objective (ablation) | Deep3D Stabilizer · NUS | cropping ratio: 0.57 -> 0.66 | Helped |
| Add geometric loss to the objective that already includes photometric loss (ablation) | Deep3D Stabilizer · NUS | cropping ratio: 0.62 -> 0.66 | Helped |
| Use depth-aware 3D reprojection rather than 2D stabilization transformations | Deep3D Stabilizer · NUS large parallax | stability: 16.68 -> 16.81 | Helped |
| Apply loss-guided spatially varying depth smoothing to dynamic-object regions | Deep3D Stabilizer · videos with dynamic objects | visual distortion | Helped |
Closest papers by meaning
- AC3D: Analyzing and Improving 3D Camera Control in Video Diffusion Transformers (CVPR 2025 · similarity 0.87)
- Harnessing Meta-Learning for Improving Full-Frame Video Stabilization (CVPR 2024 · similarity 0.86)
- CH3Depth: Efficient and Flexible Depth Foundation Model with Flow Matching (CVPR 2025 · similarity 0.85)
- Align3R: Aligned Monocular Depth Estimation for Dynamic Videos (CVPR 2025 · similarity 0.84)
- Scalable Benchmarking and Robust Learning for Noise-Free Ego-Motion and 3D Reconstruction from Noisy Video (ICLR 2025 · similarity 0.84)
- Training-free Camera Control for Video Generation (ICLR 2025 · 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.