search

yangli18 / VLTVG

Improving Visual Grounding with Visual-Linguistic Verification and Iterative Reasoning, CVPR 2022
91 stars 8 forks source link

The dataset of ReferItGame and Flickr30k is unavailable. #2

Closed WujiangXu closed 2 years ago

WujiangXu commented 2 years ago

Thank u for u solid work. I follow your repository and meet some problems.

On one hand, the link of the dataset ReferItGame in the download_data.sh script is unavailable. I can't download the file from this link. On the other hand, it is no provided way to download the dataset Flickr30k.

Can u give some suggestions?

WujiangXu commented 2 years ago

By the way, are there some interpretations for the config py file?

yangli18 commented 2 years ago

@WujiangXu Hi, the download link for ReferItGame seems to be broken. As an alternative, you can download all the images from the ImageCLEF, from which the ReferItGame is collected. Or you can try using the subset of images from here (provided by https://github.com/lichengunc/refer#prepare-images).

The images of Flickr30k need to be downloaded from its homepage. Please note that the images should be prepared according to the folder structure mentioned in get_started.md.

yangli18 commented 2 years ago

@WujiangXu The configuration files contain the arguments that we want to pass to the training/testing script. For example, "epochs=90" in the configuration file means training for 90 epochs, which is equivalent to specifying "--epoch 90" in the command line.