Yida Wang

好，这个方法有相关的参考文献说为什么适用于这个问题吗？给个链接，应该是zero padding带来的一些负面影响吧

> do you have suncg dataset？ I think they have certain legal problems about releasing the SUNCG dataset publicly now.

Files are generarted with our script from the original [SUNCG](https://sscnet.cs.princeton.edu/) binary data, although their website is currently down. The original data looks like : **NYU**: NYUtest/ NYUtrain/ **SUNCG** : SUNCGtest_49700_49884/...

> Hey, Yida Wang. I am interested in your research work and wanted to extend it. Can you please guide me on how to execute your code. I would be...

@hzxie Hi I'm wondering how you incoorporate gridding_loss in training? I have not found it in the script Thanks

@hzxie Thank you, I tried it out and the result seems to be fitting with the expected trends. Thanks for your inspiring work;)

Thanks for your explanation, this helps me a lot. Actually, your project is already very useful.

@bhack As suggested, I moved basic parser to io folder and make the parser example a standalone one in the examples/tensorflor_converter folder. This is complete in the parser stage.

Hi Yuanyuan, we tested on SUNCG and ScanNet dataset, but the completion is not satisfying compared to object completion. If you wanna have a try on scenes, then you can...

Oh wow, it is the other way out for my scenario, need to adjust the variables to double from float. Haven’t checked in details for such error, anyone have insight?