单层感知器 [英] Single Layer Perceptron

问题描述

我需要在Opencv中为面部图像实现单层感知器算法.我有训练的高斯价值.
有人可以帮忙吗?

I need to implement single layer perceptron algorithm for facial imagesin Opencv. I have the gaussian values for training .
Can someone help with how to proceed ?

推荐答案

尝试Google搜索，周围有很多文章和示例.

Try a Google search, there are lots of articles and samples around.

首先，对于大多数人来说，您是否想使用单层感知器进行面部识别?对于此类基于分类的应用程序，单层感知器并不是一个好主意.为此，您需要提出适当的功能，例如gabor系数(我不知道高斯值是否合适)，计算这些功能并使用反向传播算法训练多层感知器.

请参阅 http://research.microsoft.com/en-us/um/people /zhang/papers/ijprai.pdf [ ^ ]

至于OpenCv http://opencv.willowgarage.com/documentation/cpp/neural_networks.html [ ^ ]请参阅这些文档，以帮助您实现您的神经网络.

我以为您想要面部识别，而不是"openCV中的面部图像".

很高兴为您提供帮助:)

First and for most, do you want to do facial recognition with single layer perceptron? single layer perceptrons are not a good idea for such classification based applications. And you need to come up with appropriate features such as gabor coefficients(I don''t know if gaussian values are okay) for that,compute these features and train a multi-layer perceptron using back -propagation agorithm.

see http://research.microsoft.com/en-us/um/people/zhang/papers/ijprai.pdf[^]

As for OpenCv http://opencv.willowgarage.com/documentation/cpp/neural_networks.html[^] see these documentations to help you with the implementation of the your neural network.

I assumed you wanted facial recognition and not "facial image in openCV".

Glad to help :)

这篇关于单层感知器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IT屋！