SCB-dataset icon indicating copy to clipboard operation
SCB-dataset copied to clipboard

数据集注释的含义

Open jpyjpr opened this issue 2 years ago • 5 comments

你好,我下载了您提供的数据集,但是对于数据集中的标注文件信息有些疑惑 SCB-Dataset3数据集中0.355k_university_yolo_Dataset中的txt 文件,1-5的 标签分别指什么?,后面的是坐标值吗? 分别是什么含义呢?

jpyjpr avatar Dec 01 '23 05:12 jpyjpr

The annotation information of the data is explained in the paper

Whiffe avatar Dec 01 '23 06:12 Whiffe

好的! 感谢!!

jpyjpr avatar Dec 01 '23 06:12 jpyjpr

你好 注释中标签的含义我知道了,但是 后面坐标的含义是什么呢?我没有在论文中找到相关信息 3 0.26022567447916667 0.5867417453703703 0.08568406770833333 0.13540169444444444 1 0.10234480208333334 0.38434396296296297 0.0412551875 0.07475281481481481 我按照 左上角的坐标以及宽高(x,y,w,h)来计算 发现好像不是很正确 1701420766284

jpyjpr avatar Dec 01 '23 08:12 jpyjpr

Please lay the groundwork before you ask a question.

Whiffe avatar Dec 01 '23 12:12 Whiffe

Please lay the groundwork before you ask a question.

Dear sir, I really appreciate your work and contribution for this topic. I am doing my university research and your work helps me a lot. However, I have one question about the dataset. I read all the papers attached in the README, but I cannot find the part which explains the labels, where the txt files are saved.

For example, a line in a txt file contains as follow: 0 0.043302886740331496 0.5993243265086207 0.08140941988950276 0.19864865301724136 I know the 0 at the first of the line is the label, but how about the other four floats after that? How did you calculate them and how can I use these labels for model fine-tuning?

I hope this question can reach to you. Thank you so much!

trananh91 avatar Aug 04 '24 04:08 trananh91