We describe the 2017 version of Microsoft's conversational speech recognition
system, in which we update our 2016 system with recent developments in
neural-network-based acoustic and language modeling to further advance the
state of the art on the Switchboard speech recognition task. The system adds a
CNN-BLSTM acoustic model to the set of model architectures we combined
previously, and includes character-based and dialog session aware LSTM language
models in rescoring. For system combination we adopt a two-stage approach,
whereby subsets of acoustic models are first combined at the senone/frame
level, followed by a word-level voting via confusion networks. We also added a
confusion network rescoring step after system combination. The resulting system
yields a 5.1\% word error rate on the 2000 Switchboard evaluation set.

本文介绍了微软 2017 年的对话语音识别系统，采用了基于神经网络的声学建模和语言建模的最新进展，加入了 CNN-BLSTM 声学模型和基于字符和对话会话感知的 LSTM 语言模型进行重构，采用两阶段方法进行系统合并，并在合并后通过混淆网络进行单词级投票。该系统在 Switchboard speech recognition 数据集上的评测结果显示，词错误率为 5.1％。

微软 2017 年会话式语音识别系统

The Microsoft 2017 Conversational Speech Recognition System

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 error rate on the widely used NIST 2000 test set, and find
that our latest automated system has reached human parity. The error rate of
professional transcribers is 5.9% for the Switchboard portion of the data, in
which newly acquainted pairs of people discuss an assigned topic, and 11.3% for
the CallHome portion where friends and family members have open-ended
conversations. In both cases, our automated system establishes a new state of
the art, and edges past the human benchmark, achieving error rates of 5.8% and
11.0%, respectively. The key to our system's performance is the use of various
convolutional and LSTM acoustic model architectures, combined with a novel
spatial smoothing method and lattice-free MMI acoustic training, multiple
recurrent neural network language modeling approaches, and a systematic use of
system combination.

本篇论文测试了 Convolutional 和 LSTM acoustic model architectures 在对话语音识别上的效果，并采用 novel spatial smoothing method 和 lattice-free MMI acoustic training 方法，还有 multiple recurrent neural network language modeling approaches 以及 system combination 等方法，实现了与人类专业转录员相媲美的识别准确率。