Ran Cheng

Add support for multiple scan overlay

1

Thanks for your awesome labeling tool! As we can see that label for single scan might be in-efficient in terms of the large scale data, do you have plan to...

wrapper for kitti dataset.

Kitti Validation Results overfitting?

2

Hi, Thanks for your contributions, as we have made our own branch to train it in Kitti ([# 25 pull request](https://github.com/XinJCheng/CSPN/pull/25)) and after one week training, we got the following...

add docker

1

add docker and support remote debug with Pangolin: 

do you have plan to release the corresponding pdf?

1

as shown in the title.

I have asked the same question in DSO issue [#212](https://github.com/JakobEngel/dso/issues/212) https://github.com/tum-vision/LDSO/blob/1c1e3c90c68c48e904607ce36db6eac134852190/src/frontend/CoarseInitializer.cc#L340-L352 why note update `EAlpha` and use `EAlpha.A` to calculate `AlphaEnergy`?

Hi, Thanks for your implementation, it helped us a lot! I have further wrapped it up with a kitti loader and fixed some bugs. Best, Ran

why EAlpha not updated?

2

Recently, I carefully reviewed your code and found the following code very confusing, could you please explain why you accumulate `E` rather than `EAlpha`: https://github.com/JakobEngel/dso/blob/ae1d0b3d972367b0a384e587852bedcf5f414e69/src/FullSystem/CoarseInitializer.cpp#L502-L519 During the accumulation of energy...

resolve python setup.py install error

1

resolve the fatal error that can not find `mkl.h` error when building bdist_wheel using `setup.py`