Original article can be found here (source): Deep Learning on Medium
ML Paper Challenge Day 13, 14 — Deep Neural Networks for Acoustic Modelling in Speech Recognition
Day 13 — 14: 2020.04.24 — 25
Paper: Deep Neural Networks for Acoustic Modelling in Speech Recognition
Category: Model/Deep Learning/Speech Recognition
This is a paper illustrating the “recent development” (in year 2012) of using deep neural network to replace GMM in speech recognition.
Its idea still pre-trained with layers of RBMs, and then combine them into deep belief network and finally a pre-trained DBN-DNN, to make deep learning possible, like the early development of deep learning in image recognition.