对大词汇量汉语连续语音识别的实验结果表明:高斯模糊聚类使高斯数减少25%时,识别率提高了0.15%。
The experimental results on large vocabulary continuous Mandarin speech recognition show when the number of Gaussians is reduced by 25%, the recognition accuracy increases by 0.15%.
DRNN聚类学习的性能使得它非常适用于与在线学习方式相结合的实时语音识别系统。
The property of clustering learning of the DRNN makes it very suitable for real-time speech recognition with on-line learning ability.
指出了模糊聚类方法影响语音识别的正确率。
It is pointed that the clustering methods influence on the correct rate in speech recognition.
实验结果表明改进的算法能够很好的对汉字和语音进行聚类,改善了网络训练效果。
The improved algorithm proved good, for the characters and the phonemes was clustered well and the training effects were observed improved.
实验结果表明改进的算法能够很好的对汉字和语音进行聚类,改善了网络训练效果。
The improved algorithm proved good, for the characters and the phonemes was clustered well and the training effects were observed improved.
应用推荐