Disclosed are an apparatus and a method for detecting a speech endpoint using a WFST.
The apparatus in accordance with an embodiment of the present invention includes: a …

In this paper, a large-scale evaluation of open-source speech recognition toolkits is
described. Specifically, HTK in association with the decoders HDecode and Julius, CMU …

Disclosed embodiments include methods and systems for symbolic correction in human-
machine interfaces that comprise (a) implementing a language model;(b) implementing a …

Despite the tremendous empirical success of neural models in natural language processing,
many of them lack the strong intuitions that accompany classical machine learning …

We analyze a simple hierarchical architecture consisting of two multilayer perceptron (MLP)
classifiers in tandem to estimate the phonetic class conditional probabilities. In this …

Connectionist temporal classification (CTC) has recently shown improved performance and
efficiency in automatic speech recognition. One popular decoding implementation is to use a …

One major research challenge in the domain of the analysis of meeting room data is the
automatic transcription of what is spoken during meetings, a task which has gained …

We present the AMI 2006 system for the transcription of speech in meetings. The system was
jointly developed by multiple sites on the basis of the 2005 system for participation in the …

Using phone posterior probabilities has been increasingly explored for improving automatic
speech recognition (ASR) systems. In this paper, we propose two approaches for …

Finite-state methods are the most efficient mechanisms for analysing textual and symbolic
data, providing elegant solutions for an immense number of practical problems in …