zengyan-97 / X-VLM

X-VLM: Multi-Grained Vision Language Pre-Training (ICML 2022)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Script to generate RegionTextJsonDataset?

daizuozhuo opened this issue · comments

HI, I understand that the data cannot redistributed, but could you share the code to generate RegionTextJsonDataset from the official COCO, VG datasets so we can follow the pretraining method?

Hi,

I found that some other methods just released their data.
So, I will release the processed json files (image not included) in this week.
Please follow up then.

Is the json file avaiable now?

yes, I am also interested! :)

Hi,

The json files have been available for a while. Please see README for details.
On the other hand, we only applied some preprocessing to filter invalid bboxes in the public data. You can download the data from the original websites, and do the filtering by yourself.