Research map / 3D vision / Depth and stereo
Unsupervised Learning of Stereo Matching
Chao Zhou, Hong Zhang, Xiaoyong Shen, Jiaya Jia
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 the learned joint filtering cost aggregation branch (ablation) | stereo network · KITTI 2015 supervised | >3 pixel error: 9.23 -> 6.81 | Improved |
| add the learned joint filtering cost aggregation branch (ablation) | stereo network · KITTI 2015 supervised | >5 pixel error: 7.12 -> 5.01 | Improved |
| use learned joint filtering instead of box filter (ablation) | stereo network · KITTI 2015 supervised | >3 pixel error: 7.33 -> 6.81 | Improved |
| use learned joint filtering instead of guided filter (ablation) | stereo network · KITTI 2015 supervised | >3 pixel error: 7.13 -> 6.81 | Improved |
| use learned joint filtering instead of guided filter (ablation) | stereo network · KITTI 2015 supervised | >5 pixel error: 5.28 -> 5.01 | Improved |
| initialize with DeepMatching instead of random network (ablation) | unsupervised stereo · KITTI 2015 | >3 pixel error: 7.81 -> 7.69 | No clear effect |
| initialize with DeepMatching instead of random network (ablation) | unsupervised stereo · KITTI 2015 | >5 pixel error: 6.01 -> 5.67 | No clear effect |
| iterative confident-match unsupervised training instead of unsupervised optical-flow losses (USCNN) | unsupervised stereo · KITTI 2015 NOC test | >3 pixel error: 11.17 -> 8.61 | Improved |
| iterative confident-match unsupervised training instead of Yu et al. | unsupervised stereo · KITTI 2015 All test | >3 pixel error: 19.14 -> 9.91 | Improved |
| new network architecture instead of DispNet (supervised) | stereo network · KITTI 2015 NOC | >3 pixel error: 7.19 -> 6.81 | Improved |
| new network architecture instead of DispNet (supervised) | stereo network · KITTI 2015 | runtime: 0.12 -> 0.39 | Hurt |
Closest papers by meaning
- Stereo Matching by Training a Convolutional Neural Network to Compare Image Patches (JMLR 2016 · similarity 0.95)
- Stereo Matching Using Multi-Level Cost Volume and Multi-Scale Feature Constancy (TPAMI 2019 · similarity 0.94)
- Deep Stereo Matching With Explicit Cost Aggregation Sub-Architecture (AAAI 2018 · similarity 0.93)
- Efficient Deep Learning for Stereo Matching (CVPR 2016 · similarity 0.92)
- Revisiting Non-Parametric Matching Cost Volumes for Robust and Generalizable Stereo Matching (NeurIPS 2022 · similarity 0.91)
- Stereo Risk: A Continuous Modeling Approach to Stereo Matching (ICML 2024 · similarity 0.90)
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.
