Over the past decade or so, several advances have been made to the design of modern
large vocabulary continuous speech recognition (LVCSR) systems to the point where their …

We describe the latest version of Microsoft's conversational speech recognition system for
the Switchboard and CallHome domains. The system adds a CNN-BLSTM acoustic model to …

In this paper we describe a method to perform sequencediscriminative training of neural
network acoustic models without the need for frame-level cross-entropy pre-training. We use …

Conversational speech recognition has served as a flagship speech recognition task since
the release of the Switchboard corpus in the 1990s. In this paper, we measure the human …

Conversational speech recognition has served as a flagship speech recognition task since
the release of the Switchboard corpus in the 1990s. In this paper, we measure a human …

We present our work on end-to-end training of acoustic models using the lattice-free
maximum mutual information (LF-MMI) objective function in the context of hidden Markov …

CHiME-3 is a research community challenge organised in 2015 to evaluate speech
recognition systems for mobile multi-microphone devices used in noisy daily environments …

BACKGROUND As is well known, a language model is used to represent the language that
an automatic speech recognition (ASR) system is intended to recognize or decode. One of …

Training neural network acoustic models with sequencediscriminative criteria, such as state-
level minimum Bayes risk (sMBR), been shown to produce large improvements in …

Recent work in automatic recognition of conversational telephone speech (CTS) has
achieved accuracy levels comparable to human transcribers, although there is some debate …