利用SVM进行语音情感识别

利用的是c++语音编写的语音情感识别的代码，样本为提取的特征文件，一个样本具有39个特征，还有类别文件用数字表示，利用了SVM的好多类型，采用了核函数的很多类型，调试成功即可用，只是识别率略低25%，只是作为参考，大家可以自己改进。其实是从别的地方下载的，感谢作者！

EmotionClassifierBySvm
Debug
ca.dsp
ca.dsw
ca.ncb
ca.opt
ca.plg
class.txt
class120.txt
class160.txt
class200.txt
class240.txt
class280.txt
class320.txt
class360.txt
class40.txt
class400.txt
class440.txt
class554.txt
class80.txt
c_svc_svr.cpp
c_svc_svr.h
feature120.txt
feature160.txt
feature200.txt
feature240.txt
feature280.txt
feature320.txt
feature360.txt
feature40.txt
feature400.txt
feature440.txt
feature554.txt
feature80.txt
final.txt
m.cpp
model.txt
output.txt
predictClass1.txt
predictClass40.txt
PredictFeature1.txt
PredictFeature40.txt
SampleNum_RecognitionRate.txt
svm.cpp
svm.h
test.cpp
test_g_c_RecognitionRate.txt
~$实验结果.doc
~WRL0005.tmp
~WRL0593.tmp
~WRL3596.tmp