Min Wang

Hi Sylvain, Since the experiments and blog are from 2018, is that still the fast way to train networks now? How do you choose the fixed epoch of one circle...

architecture is different from the model in paper (up sampling stage and fusion module)

9

Hi Yisheng,  The network 's up sampling stage is not correct. The 3rd layer of up sampling and the conv2d afterwards are swapped, i.e., the 3rd layer of up...

Adaptation to New Dataset

1

Hi, Could you write a basic instruction for preparation for a new Dataset?

lower mIoU than the result of the paper

5

Hi, I used the .yml file to create conda environment and default setting to train NUYv2 and the result 47. 48 is much lower than the result of the paper...

Cannot dlopen some GPU libraries

5

 Hi all, I followed ReadMe for installation but I got lib errors when I train models via train.py. I have checked tf-GPU, cuda and cudnn versions. Could you help...

Hi Matthew, Could you add CADC data loader to MMdetection3d? Thanks! Best wishes, Min

HI, does anyone know the meaning and format of trainval txt? It seems that it contain all the images... What's trainval? train and evaluation? 

Hi, According to the paper: Random Sampling (RS): We implement random sam- pling with the python numpy package. Specifically, we first use the numpy function numpy.random.choice() to generate K indices....

Here is my user setting JSON: ``` { "workbench.colorTheme": "Default Dark+", "terminal.integrated.inheritEnv": false, "remote.SSH.connectTimeout": 60, "security.workspace.trust.untrustedFiles": "open", "editor.accessibilitySupport": "on", "files.autoSave": "afterDelay", "editor.fontSize": 24, "[cpp]": { "editor.defaultFormatter": "xaver.clang-format" }, "clang-format.fallbackStyle": "Google",...