![](https://img-blog.csdnimg.cn/20201014180756918.png?x-oss-process=image/resize,m_fixed,h_64,w_64)
计算机视觉
lizhaohu01
这个作者很懒,什么都没留下…
展开
-
Image-based localization using LSTMs for structured feature correlation
文中提出很多场景使用deeplearning的方式还没有传统的sift的效果好 本文主要是解决两种场景:1:无纹理。2:重复纹理 也是一个end-to-end的直接回归得到6dof pose的文章,只不过不仅仅使用了CNN。文章是CNN+LSTM的方式。仍然使用pose 误差:整体网路框架:使用预训练的网络googlenet,同posenet一样最后的fc层输出的是2048维度的global feature。文章指出2048维度的feature可以表达的数据...原创 2020-06-11 15:47:07 · 529 阅读 · 0 评论 -
Learning 2D–3D Correspondences To Solve The Blind Perspective-n-Point Problem
When correspondences are known, the problem reduces to the standard PnP problem [10,17,33,19]. When correspondences are unknown, the problem is blind PnP, for which several traditional geometry-based methods were proposed, in- cluding SoftPoSIT [7], Blin..原创 2020-06-05 14:43:52 · 809 阅读 · 0 评论