论文标题
在语音识别方面公平：发现和缓解绩效差异
Toward Fairness in Speech Recognition: Discovery and mitigation of performance disparities
论文作者
论文摘要
至于其他形式的AI，最近已经对不同用户同类的性能差异进行了研究。在语音识别方面实现公平性的一种方法是（1）确定遭受低标准表现的说话者队列，以及（2）采取针对发现同类的公平性缓解措施。在本文中，我们使用产品规模AI助手语音识别系统的数据进行了发现和缓解性能差异的初步发现。我们将基于地理和人口统计学信息的队列发现与一种更可扩展的方法进行比较，该方法将使用扬声器嵌入技术分组没有人类标签的说话者。为了进行公平性缓解，我们发现通过其他输入变量对代表性不足的队列的过度采样不足，以及对扬声器队列的建模，从而降低了表现和底部表现的人群之间的差距，而不会降低整体识别精度。
As for other forms of AI, speech recognition has recently been examined with respect to performance disparities across different user cohorts. One approach to achieve fairness in speech recognition is to (1) identify speaker cohorts that suffer from subpar performance and (2) apply fairness mitigation measures targeting the cohorts discovered. In this paper, we report on initial findings with both discovery and mitigation of performance disparities using data from a product-scale AI assistant speech recognition system. We compare cohort discovery based on geographic and demographic information to a more scalable method that groups speakers without human labels, using speaker embedding technology. For fairness mitigation, we find that oversampling of underrepresented cohorts, as well as modeling speaker cohort membership by additional input variables, reduces the gap between top- and bottom-performing cohorts, without deteriorating overall recognition accuracy.
