-
The algorithm section needs expansion as a key point of research.
1. Improve the framework and API to allow different algorithms to be specified and evaluated. What is the best framework - is there…
-
或者SLAM十四讲
-
hello,can you give me some suggestions on why this method is so insensitive to vehicles?
![Uploading Screenshot from 2020-05-26 22-07-46.png…]()
-
## 🚀 Feature
@jatentaki reported that kornia functions are clean, but slower that optimized ones.
I have quickly checked and yes - for example, `batch_episym` from OANet (https://github.com/zjhthu…
-
## 🐛 Bug
The [docs](https://kornia.readthedocs.io/en/v0.5.0/geometry.epipolar.html#kornia.geometry.epipolar.projections_from_fundamental) state that the shape of input `F` is `(*, 3, 3)`, but in pr…
-
原文：[Yin的笔记本，非常推荐](http://www.yindaheng98.top/%E5%9B%BE%E5%BD%A2%E5%AD%A6/%E5%8F%8C%E7%9B%AE%E7%9B%B8%E6%9C%BA.html#%E4%B8%89%E8%A7%92%E5%AE%9A%E4%BD%8D-triangulation)
图片部分来自[知乎，大黑](https://zhuanlan.z…
-
hi @Kai-46 ,
thank you for your excellent job!! However I encountered a problem when trying to training using my images. Acturally, I rendered images using blender renderer, and keep the training set…
-
-
### Describe the bug
When calling kornia.geometry.epipolar.essential.run_5point I get a segmentation fault, running into the following error:
`Intel MKL ERROR: Parameter 3 was incorrect on entry t…
-
hello, in the code **estimatepose.py** ,only two poses of img1 and img2 are estimated, but in your report, consecutive images can be estimate poses. how to do that?
thanks in advance.