Newbeeer / L_DMI

Code for NeurIPS 2019 Paper, "L_DMI: An Information-theoretic Noise-robust Loss Function"

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

clothing1M

xjtushujun opened this issue · comments

   Can you provide the following files: noisy_train.txt,clean_val.txt,clean_test.txt.
    if train==True:
        flist = os.path.join(root, "annotations/noisy_train.txt")
    if valid==True:
        flist = os.path.join(root, "annotations/clean_val.txt")
    if test==True:
        flist = os.path.join(root, "annotations/clean_test.txt")

Hi, sorry for the late response. I have uploaded these files. BTW do you have the image data for Clothing1M? If not I can upload it on another website.

Thanks, I have download the original data. However, the name of the .txt files in the original data is different from yours. Why?

We preprocess three original files (clean_train_key_list.txt, clean_val_key_list.txt, noisy_train_key_list.txt). The indices in the original file are consistent with the image path in the current .txt files. We combine the original indices files and label files into one .txt file.

Hi! It seems that there is a small typo in file clean_train.txt:
line 10620: images/2/26/3761328114,905362226.jpg don't be followed by the corresponding label.