In this paper, we present a framework to evaluate the human corrections of a speaker diarization system. We propose four elementary actions to correct the diarization (“Create a boundary”, “Delete a boundary”, “Create a speaker label” and “Change the speaker label”) and we propose an automaton to simulate the correction sequence. A metric is described to evaluate the correction cost. The framework is evaluated using French broadcast news drawn from the following campaigns: REPERE, ESTER and ETAPE.

中文翻译：

---

在计算机辅助的说话人区分系统中评估人工校正

在本文中，我们提出了一个评估说话人差异化系统的人为校正的框架。我们提出了四个基本动作来纠正差异化（“创建边界”，“删除边界”，“创建说话者标签”和“更改说话者标签”），并提出一个自动机来模拟校正序列。描述了用于评估校正成本的度量。使用从以下活动中提取的法语广播新闻对框架进行评估：REPERE，ESTER和ETAPE。