search

AnantLabs / facerecog

Automatically exported from code.google.com/p/facerecog
0 stars 0 forks source link

能否解释一下这个现象 #4

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
您好：
我运行了这个工程，注册了两张image，A和B，我识别A的时候，
却找到了B，也就是说A和B计算出得那个向量距离0.9914....比A和
其本身计算出来的距离0.9811...还要小。

这是为什么呢？为何一张图片与其自身得到的向量距离不是��
�小的？请不吝赐教。

谢谢

Original issue reported on code.google.com by linu...@gmail.com on 12 Oct 2011 at 2:57

GoogleCodeExporter commented 9 years ago 
I think this is strange. The normalized inner-product distance between an image 
and itself should always be 1, because the pre-processing and feature 
extraction algorithm don't change. Can you try to output the normalized face 
images and see if they are the same between the enrolled image A and the probe 
image A?

Original comment by xje...@gmail.com on 26 Nov 2011 at 12:28