本文介绍了有没有办法在张量流中向conv2d添加规范化的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!
问题描述
为了正确进行模式匹配,卷积需要规范化
in order to do pattern matching properly convolutions require normalizationhttps://en.wikipedia.org/wiki/Cross-correlation#Normalized_cross-correlation
不幸的是,我找不到一种方法来使输入标准化
unfortunately I can't find a way how to make input normalization for conv2d function.
是否隐藏在实现中?
推荐答案
原来,我正在寻找tf.local_response_normalization()出于某种奇怪的原因,它出现在conv2层之后,并且在示例中很少使用
It turned out, I was looking for tf.local_response_normalization (https://www.tensorflow.org/versions/r0.11/api_docs/python/nn/normalization) for some strange reason it goes after conv2 layers and used not very often in examples
这篇关于有没有办法在张量流中向conv2d添加规范化的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!