A hierarchical structure for feed-forward neural networks and its application to speaker-independent speech recognition
Xu Dongxin, Taiyi Huang, Lin Zhiwei · 2002
A hierarchical structure for feed-forward neural networks which is a generalization of the time delay neural network (TDNN) model is proposed. A comparative experiment has been done on speaker-independent speech recognition. A slightly better performance, compared with the TDNN model, is achieved.>