欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
研究了使用卷积神经网络构造模式分类器，并用于连续语音识别的研究。CNNs相比于广泛使用于语音识别中的深层神经网络（Deep Neural Network，DNNs），能在保证性能的同时，大大压缩模型的尺寸。在标准语音识别库TIMIT上的实验结果证明，相比传统DNN模型，CNN模型的识别性能更好，同时其模型规模和计算量都有明显降低。
英文摘要：
Convolutional Neural Networks （CNNs） are investigated for continuous speech recognitions in the paper. Compared to Deep Neural Networks （DNNs） , which have been proven to be successful in many speech recognitio.n tasks nowadays, CNNs can reduce the NN model sizes significantly, and at the same time achieve even better recognition accuracies. Experiments on standard speech corpus TIMIT showed that CNNs outperformed DNNs in accuracy.
同期刊论文项目
同项目期刊论文
A Comparative Study on Selecting Acoustic Modeling Units in Deep Neural Networks based Large Vocabul
A Novel Multichannel Audio Signal Compressing Method Based on Tensor Representation and Decompositio
Context-based adaptive arithmetic coding in time and frequency domain for the lossless compression o
Objective and Subjective Investigation on a Novel Method for Digital Reverberator Parameters Estimat
期刊信息
位置：