Unsupervised Learning of Stereo Matching
Cost-volume Computation 用对应的分支来计算cost-volume，输入即左右图像，生成cost-volume。这部分是由八个卷基层构成的双塔结构。每个层后面有normalization和Relu层。这些层对两个图像的每个块都会产生特征向量。这些特征向量再进入correlation layer，计算得到cost-volume，这个correlation层就是用的Dispnet
相关文章
- 1. Self-Supervised Learning for Stereo Matching with Self-Improving Ability
- 2. 《Efficient Deep Learning for Stereo Matching》学习笔记
- 3. Pyramid Stereo Matching Network
- 4. 【论文笔记-Stereo】Pyramid Stereo Matching Network
- 5. 论文《Patchmatch-Based Robust Stereo Matching Patchmatch-Based Robust Stereo Matching》学习
- 6. Deep Temporal Clustering: Fully unsupervised learning of time-domain features
- 7. GoodPoint: unsupervised learning of keypoint detection and description∗
- 8. Unsupervised Learning of Edges 论文阅读
- 9. [解读] Deep Clustering for Unsupervised Learning of Visual Features
- 10. Dynamic Label Graph Matching for Unsupervised Video Re-Identification
- 更多相关文章...
- • XSLT
- • XSLT
- • Java Agent入门实战（一）-Instrumentation介绍与使用
- • Java Agent入门实战（三）-JVM Attach原理与使用
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. Self-Supervised Learning for Stereo Matching with Self-Improving Ability
- 2. 《Efficient Deep Learning for Stereo Matching》学习笔记
- 3. Pyramid Stereo Matching Network
- 4. 【论文笔记-Stereo】Pyramid Stereo Matching Network
- 5. 论文《Patchmatch-Based Robust Stereo Matching Patchmatch-Based Robust Stereo Matching》学习
- 6. Deep Temporal Clustering: Fully unsupervised learning of time-domain features
- 7. GoodPoint: unsupervised learning of keypoint detection and description∗
- 8. Unsupervised Learning of Edges 论文阅读
- 9. [解读] Deep Clustering for Unsupervised Learning of Visual Features
- 10. Dynamic Label Graph Matching for Unsupervised Video Re-Identification