Self-Supervised Learning for Stereo Matching with Self-Improving Ability
首先呢，这是一个非监督算法，因此它约束的方式就是左右一致性检测，用warp来处理左右图，详见3.1。作者声情并茂的讲述自己就是不要gt。。 网络结构 五部分组成 特征提取 交叉特征向量融合 3D特征匹配 soft-argmin 最后通过图像warp来做约束。 特征提取 有了特征提取，就可以不虚那些复杂区域啦。以往传统的方法相当于是取像素域的原始特征，而深度学习取得特征则是自己去学的 具体特征怎么取
相关文章
- 1. Unsupervised Learning of Stereo Matching
- 2. 《Efficient Deep Learning for Stereo Matching》学习笔记
- 3. Improved Stereo Matching with Constant Highway Networks and Reflective Confidence Learning
- 4. 论文阅读《Cascade Residual Learning: A Two-stage Convolutional Neural Network for Stereo Matching》
- 5. Confidence Inference for Focused Learning in Stereo Matching 论文笔记
- 6. Deep Stereo Matching with Explicit Cost Aggregation Sub-Architecture
- 7. Computing the Stereo Matching Cost with a Convolutional Neural Network
- 8. Segment-based Disparity Refinement with Occlusion Handling for Stereo Matching翻译和理解
- 9. Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching(CVPR 2020)
- 10. Pyramid Stereo Matching Network
- 更多相关文章...
- • Swift for 循环 - Swift 教程
- • Scala for循环 - Scala教程
- • Java Agent入门实战（三）-JVM Attach原理与使用
- • 为了进字节跳动，我精选了29道Java经典算法题，带详细讲解
相关标签/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
欢迎关注本站公众号,获取更多信息
相关文章
- 1. Unsupervised Learning of Stereo Matching
- 2. 《Efficient Deep Learning for Stereo Matching》学习笔记
- 3. Improved Stereo Matching with Constant Highway Networks and Reflective Confidence Learning
- 4. 论文阅读《Cascade Residual Learning: A Two-stage Convolutional Neural Network for Stereo Matching》
- 5. Confidence Inference for Focused Learning in Stereo Matching 论文笔记
- 6. Deep Stereo Matching with Explicit Cost Aggregation Sub-Architecture
- 7. Computing the Stereo Matching Cost with a Convolutional Neural Network
- 8. Segment-based Disparity Refinement with Occlusion Handling for Stereo Matching翻译和理解
- 9. Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching(CVPR 2020)
- 10. Pyramid Stereo Matching Network