Research map / 3D vision / Depth and stereo
Single View Stereo Matching
Read the paperCheck 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 |
|---|---|---|---|
| Decompose monocular depth estimation into view synthesis followed by stereo matching | Single View Stereo Matching · KITTI Eigen test, 0-80 m | ARD: 0.113 -> 0.094 | Helped |
| Use the view-synthesis-plus-stereo pipeline for monocular input | Single View Stereo Matching · KITTI 2015 Stereo test | D1-all: 25.27 -> 24.44 | Helped |
| End-to-end finetune both networks with 700 samples rather than no joint finetuning (ablation) | Single View Stereo Matching · KITTI Eigen test, 0-80 m | ARD: 0.102 -> 0.094 | Helped |
| Increase joint end-to-end finetuning data from 200 to 500 samples (ablation) | Single View Stereo Matching · KITTI Eigen test, 0-80 m | ARD: 0.100 -> 0.094 | Helped |
| Increase joint end-to-end finetuning data from 500 to 700 samples (ablation) | Single View Stereo Matching · KITTI Eigen test, 0-80 m | ARD: 0.094 -> 0.094 | No clear effect |
| End-to-end finetune the view synthesis network only with 700 samples (ablation) | Single View Stereo Matching · KITTI Eigen test, 0-80 m | ARD: 0.102 -> 0.099 | Helped |
| End-to-end finetune the stereo matching network only with 200 samples (ablation) | Single View Stereo Matching · KITTI Eigen test with true right images, 0-80 m | ARD: 0.062 -> 0.065 | Hurt |
| End-to-end finetune the stereo matching network only with 700 samples (ablation) | Single View Stereo Matching · KITTI Eigen test with true right images, 0-80 m | ARD: 0.062 -> 0.053 | Helped |
| Use stereo matching on synthesized views instead of the primitive probabilistic disparity (ablation) | Single View Stereo Matching · KITTI Eigen test, 0-80 m | ARD: 0.212 -> 0.094 | Helped |
| Remove 200 high-quality KITTI disparity labels from stereo-network training (ablation) | Single View Stereo Matching · KITTI Eigen test, 0-80 m | ARD: 0.094 -> 0.119 | Hurt |
| Finetune the model without the 200 high-quality KITTI labels (ablation) | Single View Stereo Matching · KITTI Eigen test, 0-80 m | ARD: 0.119 -> 0.101 | Helped |
Closest papers by meaning
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.91)
- MonSter: Marry Monodepth to Stereo Unleashes Power (CVPR 2025 · similarity 0.90)
- Stereo Anywhere: Robust Zero-Shot Deep Stereo Matching Even Where Either Stereo or Mono Fail (CVPR 2025 · similarity 0.90)
- A Simple yet Universal Framework for Depth Completion (NeurIPS 2024 · similarity 0.89)
- DME: Unveiling the Bias for Better Generalized Monocular Depth Estimation (AAAI 2024 · similarity 0.89)
- Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation (CVPR 2024 · similarity 0.89)
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.