youmi-zym / acfnet Goto Github PK

Hi,

First of all great work. Really impressed with your work. But I am curious to see the lines/stripes that appear on your KITTI error maps (see below for example). May I know what could be the reason for that occurrence?

First,thanks for your outstanding work, open source spirit and the patience to the problem.Would you please push the complete code on github?Can you provide the train files?
And I also try to repo the acfnet according to the DensematchingBenchmark.
Describe the bug
When I use 'python tools/train.py configs xxx ', this error happened.
File "tools/train.py", line 55, in main
cfg = Config.fromfile(args.config)
File "/home/anaconda3/envs/dmn/lib/python3.7/site-packages/mmcv/utils/config.py", line 211, in fromfile
use_predefined_variables)
File "/home//anaconda3/envs/dmn/lib/python3.7/site-packages/mmcv/utils/config.py", line 142, in _file2dict
mod = import_module(temp_module_name)
File "/home/anaconda3/envs/dmn/lib/python3.7/importlib/init.py", line 127, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
File "", line 1006, in _gcd_import
File "", line 983, in _find_and_load
File "", line 967, in _find_and_load_unlocked
File "", line 677, in _load_unlocked
File "", line 728, in exec_module
File "", line 219, in _call_with_frames_removed
File "/tmp/tmpvmtj58zj/tmpdbc_zg73.py", line 147, in
NameError: name 'vis_data_root' is not defined
Exception ignored in: <function _TemporaryFileCloser.del at 0x7fee72522200>
Traceback (most recent call last):
File "/home/anaconda3/envs/dmn/lib/python3.7/tempfile.py", line 448, in del
self.close()
File "/home/anaconda3/envs/dmn/lib/python3.7/tempfile.py", line 444, in close
unlink(self.name)
FileNotFoundError: [Errno 2] No such file or directory: '/tmp/tmpvmtj58zj/tmpdbc_zg73.py'
Environment
python 3.7 pytorch 1.5 GCC4.6.5 cuda10.2 mmcv 0.4.4

First,thanks for your outstanding work and open source spirit.But the code currently provided seems to be incomplete.Would you please push the complete code on github?

Hi, thanks to your paper and code, it's really impressive.I have a question about the calculation of confidence map.You use 3x3 window as neighborhood to calculate confidence map.Is there any reason to choose the size of window?I think 3x3 maybe too samll, it doesn't have enough receptive field for edge or occlusion.