Big dumb neural nets: a working brute force approach to speech recognition

Neural networks with over a million connections have been trained using online backpropagation and a speech corpus with over 6 million training examples (speech frames). These networks provide phonetic probabilistic estimates that are used in a continuous speech recognizer. Issues of the network training and application are discussed in this paper.<<ETX>>

[1]  Jerome A. Feldman,et al.  A supercomputer for neural computation , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[2]  Hervé Bourlard,et al.  Connectionist probability estimators in HMM speech recognition , 1994, IEEE Trans. Speech Audio Process..