Research map / 3D vision / Depth and stereo
Bilateral Cyclic Constraint and Adaptive Regularization for Unsupervised Monocular Depth Prediction
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 the full adaptive-regularization and bilateral-cyclic-consistency objective (ablation) | KITTI 2015 stereo-unsupervised model · KITTI 2015 split | D1-all: 30.318 -> 28.142 | Helped |
| Add adaptive regularization (ablation) | Godard et al. model · KITTI 2015 split | D1-all: 30.272 -> 30.132 | Helped |
| Add bilateral cyclic consistency instead of left-right consistency (ablation) | Model with adaptive regularization and Laplacian edge weights · KITTI 2015 split | D1-all: 29.536 -> 28.142 | Helped |
| Use bidirectional Laplacian-based edge awareness rather than image-gradient edge weights (ablation) | Model with adaptive regularization and bilateral cyclic consistency · KITTI 2015 split | D1-all: 28.942 -> 28.142 | Helped |
| Use the two-branch decoder (ablation) | Ours Full Model · KITTI 2015 split | D1-all: 28.142 -> 27.149 | Helped |
| Add the two-branch decoder (ablation) | Ours Full Model · KITTI Eigen Split, 80m | AbsRel: 0.135 -> 0.133 | Helped |
| Add the two-branch decoder (ablation) | Ours Full Model · KITTI Eigen Split, 50m | AbsRel: 0.128 -> 0.126 | Helped |
Closest papers by meaning
- Single View Stereo Matching (CVPR 2018 · similarity 0.91)
- P3Depth: Monocular Depth Estimation With a Piecewise Planarity Prior (CVPR 2022 · similarity 0.91)
- Unsupervised Monocular Depth Estimation With Left-Right Consistency (CVPR 2017 · similarity 0.91)
- A Simple yet Universal Framework for Depth Completion (NeurIPS 2024 · similarity 0.89)
- MonoInstance: Enhancing Monocular Priors via Multi-view Instance Alignment for Neural Rendering and Reconstruction (CVPR 2025 · similarity 0.88)
- RefineAny3D: Depth Refinement as Semantic Alignment for Monocular 3D Detection (NeurIPS 2026 · similarity 0.88)
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.