shvdiwnkozbw / Self-supervised-Video-Concept

Code for Static and Dynamic Concepts for Self-supervised Video Representation Learning.
10 stars 1 forks source link

提问 #1

Open LijunZhang01 opened 2 years ago

LijunZhang01 commented 2 years ago

您好,您代码注释中写了“feat b x n x 512 ”,我没太明白n所代表的是什么信息,能解答一下吗?非常感谢

shvdiwnkozbw commented 2 years ago

您好,n代表是视角数量,例如在计算时包含rgb的两个视角,static frame帧两个视角,frame difference两个视角时,n为6

LijunZhang01 commented 2 years ago

您好,n代表是视角数量,例如在计算时包含rgb的两个视角,static frame帧两个视角,frame difference两个视角时,n为6

好的好的,我明白了,非常感谢您的解答