search

ShichenLiu / SoftRas

Project page of paper "Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoning"
MIT License
1.24k stars 156 forks source link

What's the order of this dataset? #27

Closed qiaogh97 closed 5 years ago

qiaogh97 commented 5 years ago

Hello, ShichenLiu ! I find there is no hashname in your datasets which are .npz files? Therefore, I want to know how to sort them in order?

ShichenLiu commented 5 years ago

Hi,

If you mean the id_to_name map, the number is the ShapeNet category id, you can find it under this site.

qiaogh97 commented 5 years ago

No, I know this id_to_name map. For example, chair category has 6778 models, I see 70% for train, 20% for test and 10% for val.

I download Shapenet datasets from https://www.shapenet.org/. But I can't associate this dataset with Shapenet datasets from official website.

Every shapenet model file have a hash name, like '1b84dededd445058e44a5473032f38f' '8ea3fa037107ec737426c116c412c132'. Do you know these .npz file's hashaname?

I sort Shapenet dataset by file name, but these npz file have no law to keep. So I want to know how original author deal this file from official dataset.

Thank you for your reply !!

ShichenLiu commented 5 years ago

Oh I see. Then you might be able to find some helpful information here :)

qiaogh97 commented 5 years ago

Yes！！That‘s I need, thank you！

zyz-notebooks commented 4 years ago

Yes！！That‘s I need, thank you！ hello, I would like to ask you a question. When I load data during training, it takes up a lot of my memory. My memory configuration is only 16g. How should I set it to avoid this situation? I look forward to your reply

qiaogh97 commented 4 years ago

You can adjust BATCH_SIZE to suitable value instead of read all the dataset once. It is convenient that using class DataLoader in pytoch.