search

lluisgomez / TextTopicNet

Self-supervised learning of visual features through embedding images into text topic spaces
95 stars 28 forks source link

Folder is missing #5

Closed irazakharchenko closed 2 years ago

irazakharchenko commented 6 years ago

Hi,

Congratulations on your work "Self-supervised learning of visual features through embedding images into text topic spaces"!

We are trying to reproduce your experiments using your GitHub directory. However, we cannot find the following files required in voc_2007_classification.py:

gt_root = '/home/yash/LDA_wikipedia_imageCLEF/wiki_large/TextTopicNet/nets/CaffeNet/SVMs/VOC2007/GT_labels/' gt_train_sufix = '_train.txt' gt_val_sufix = '_val.txt' gt_test_sufix = '_test.txt'

Can you please provide us with those files?

yash0307 commented 6 years ago

Hello,

Thanks for appreciating our work.

This folder corresponds to the PASCAL VOC 2007 dataset used for the evaluation of learned visual features on image classification task.

You can download this dataset from this link: http://host.robots.ox.ac.uk/pascal/VOC/voc2007/index.html

Once the download is done, please specify the path as per location on your device to "gt_root" variable.