search

sparkydogX / sparkydogx_blog_comment

Comments for https://sparkydogx.github.io
1 stars 0 forks source link

根据文件名从json文件中提取信息 | SparkydogX Blog #35

Open sparkydogX opened 5 years ago

sparkydogX commented 5 years ago

https://sparkydogx.github.io/2017/10/21/extract-subset/

在遇到大型数据库时，从中选取一个子集先行训练往往是一个比较好的选择。代码如下：12345678910111213141516171819import jsonimport osimgpath='/data2/guoyu/subset/'labeljsonpath='/data2/guoyu/keypoint_validation_annotations_2017091