有关RTMDet的backbone&neck参数配比的问题(Questions about the ratio of the backbone&neck parameters of RTMDet) #935
Unanswered
little-spoon
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
中文:RTMDet里面提到的 backbone {expand_ratio} 和 neck {expand_ratio} 这两个参数的和是为1的吗?那为什么论文中Table5(c)中两个值的和不为1呢?这两个值的具体含义究竟是什么呢?我如果进行不同配比的实验,具体应该如何调整这两个值才是比较合理的呢?(例如0.6/0.4,0.4/0.6,04/0.4,0.5/0.4等等,这些是否都是合理的配比?)
English:Is the sum of backbone {expand_ratio} and neck {expand_ratio} mentioned in RTMDet equal to 1? So why doesn't the sum of the two values in Table5(c) in the paper equal 1? What exactly do these two values mean? If I conduct experiments with different ratios, how should I adjust these two values is more reasonable? (e.g. 0.6/0.4, 0.4/0.6, 04/0.4, 0.5/0.4, etc., are these all reasonable ratios?)
Beta Was this translation helpful? Give feedback.
All reactions