Open jaepoong opened 4 months ago
Hi, thanks for your interest in our work, but I'm a little confused. Did you mean cgl-v2 has a total of 39,931 images without a text category.
Thank you for your message. It seems like you are comparing the number of images in CGL-v2, mentioning that there are around 60,000 images with 216 images excluded for text categories. However, I found that the text annotations you provide cover only 39,931 images. Is there any reason??
Thank you
Hello! I noticed that in the CGLdataset, there are 216 annotated images without a text category. However, when I looked at the train set of cgl-v2, it appears to have a total of 39,931 images with a text category. I want to know about the reason of the excepted category. Thank you.