It s a neural segreatory that recognize people and speech with these. If a people is not recognized Rubycon ask to create a training with thi new people for a next recognition. Rubycon use the Microsoft Speech sdk, gann neural network and other lib

RNNLIB可以用来做语音识别、手写字符识别。由大牛Alex Graves编写，专门做RNN、LSTM的研究。他的主页是http://www.cs.toronto.edu/~graves/-RNNLIB is a recurrent neural network library for sequence labelling problems, such as speech and handwriting recognition.

BP神经网络基本原理概述：这种网络模型利用误差反向传播训练算法模型，能够很好地解决多层网络中隐含层神经元连接权值系数的学习问题，它的特点是信号前向传播、误差反向传播，简称BP（Back Propagation）神经网络。BP学习算法的基本原理是梯度最快下降法，即通过调整权值使网络总误差最小，在信号前向传播阶段，输入信号经输入层处理再经隐含层处理最后传向输出层处理；在误差反向传播阶段，将输出层输出的信号值与期望输出信号值比较得到误差，若误差较大则把误差信号传回隐含层直至输入层，在各层神经元中使用