nywang16 / Pixel2Mesh

Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images. In ECCV2018.
http://openaccess.thecvf.com/content_ECCV_2018/papers/Nanyang_Wang_Pixel2Mesh_Generating_3D_ECCV_2018_paper.pdf
Apache License 2.0
1.64k stars 294 forks source link

upgrade to Python 3 #77

Open mewmew opened 4 years ago

mewmew commented 4 years ago

As Python 2 has reached the end of its life, as of the 1st of January 2020, it would be good to see Pixel2Mesh updated to support Python 3. Has there been any work in this direction?

Cheers, Robin

Deprication message from pip:

DEPRECATION: Python 2.7 reached the end of its life on January 1st, 2020. Please upgrade your Python as Python 2.7 is no longer maintained. A future version of pip will drop support for Python 2.7. More details about Python 2 support in pip, can be found at https://pip.pypa.io/en/latest/development/release-process/#python-2-support

walsvid commented 4 years ago

Hi, @mewmew . We do have plan to upgrade to python3, we are working in progress to complete the codebase. After the update, I will inform everyone.

mewmew commented 4 years ago

Hi, @mewmew . We do have plan to upgrade to python3, we are working in progress to complete the codebase. After the update, I will inform everyone.

Hi @walsvid,

Thanks for the update. That's amazing news!

Happy coding!

Cheers, Robin

5agado commented 4 years ago

I was able to run the demo in Python 3.6 and Tensorflow 1.15 with very little changes:

Fix the only print at the end of demo.py and use pickle (replace import cPickle as pickle line with import pickle), with this adjustment when loading for compatibility

pkl = pickle.load(open('Data/ellipsoid/info_ellipsoid.dat', 'rb'), encoding='latin1')
lkf1996 commented 4 years ago

我能够在Python 3.6和Tensorflow 1.15中运行该演示,而几乎没有什么变化:

修复demo.pyprint末尾的_唯一内容,_并使用pickle(用替换import cPickle as pickleimport pickle),并在加载时进行此调整以实现兼容性

pkl = pickle.load(open('Data/ellipsoid/info_ellipsoid.dat', 'rb'), encoding='latin1')

excuse me ,may I ask which version of CUDA you used? CUDA8.0?

walsvid commented 3 years ago

我能够在Python 3.6和Tensorflow 1.15中运行该演示,而几乎没有什么变化: 修复demo.pyprint末尾的_唯一内容,_并使用pickle(用替换import cPickle as pickleimport pickle),并在加载时进行此调整以实现兼容性

pkl = pickle.load(open('Data/ellipsoid/info_ellipsoid.dat', 'rb'), encoding='latin1')

excuse me ,may I ask which version of CUDA you used? CUDA8.0?

Hi, @lkf1996 Yes, CUDA 8.0 is suitable for this algorithm.