search

huanghoujing / beyond-part-models

PCB of paper: Beyond Part Models: Person Retrieval with Refined Part Pooling, using Pytorch
331 stars 81 forks source link

Preprocessing data with torchvision.transforms is much slower #37

Open vhank opened 3 years ago

vhank commented 3 years ago

I try preprocessing data with torchvision.transforms, such as ToTensor and RandomHorizontalFlip, and keeping other codes changed. I find that loading data becomes much slower. I wonder if you know the reason. Thank you!

huanghoujing commented 3 years ago

Hi, vhank. I did not try the code in your way, and I could not tell the reason for the problem you encountered. Sorry for that.

I recommend you not just change dataset preprocessing, but also change data loader into the one implemented by Pytorch.