Achieving Human Parity in Conversational Speech Recognition
Abstract
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.
中文速览
微软的语音识别系统在经典的对话语音识别基准测试（NIST 2000评测集）上首次达到了"人类同等水平"：他们先找专业转录员测出人类的真实错误率——Switchboard部分为5.9%、CallHome部分为11.3%——然后用自家系统分别跑出5.8%和11.0%，双双超越人类基准。系统的核心是多种卷积神经网络（CNN，包括VGG、ResNet、LACE）与双向长短时记忆网络（BLSTM）的组合，配合一种新颖的"空间平滑（spatial smoothing）"正则化方法来提升声学模型精度，再叠加无格网MMI声学训练、多种循环神经网络语言模型以及系统融合技术，最终形成整体优势。这项结果标志着在自发性对话语音这一语音识别最难任务上，机器首次与人类旗鼓相当，具有重要的里程碑意义，也为后续研究提供了更清晰的人类基线参照。
原文 arXiv:1610.05256;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1610.05256v2