arXiv · cs.AI · Santhoshkumar V · 10 小时前模型评分44
TamilEOT:泰米尔语电话语音语义轮次结束检测数据集与模型
TamilEOT: A Dataset and Model for Semantic End-of-Turn Detection in Tamil Telephone Speech
导读
研究者发布TamilEOT,据称是首个面向南印度语言的语义轮次结束检测开放数据集与检测器,含116通真实泰米尔语电话对话中的18,485个标注轮次边界。在30通未见通话的4,168个片段上,准确率从零样本70.30%提升至83.71%(8.7 MB)和86.13%(21 MB),ROC-AUC从0.751升至0.921,两个模型在笔记本CPU单线程下运行均低于150毫秒。
分类置信度
中 · 待人工复核
- 存在未解决的相邻类别:foundation-models,需要人工复核。
- 只有单一媒体来源,需要人工复核。
来源级别:媒体与个人
事件类型:研究发布
证据摘录
- “We release TamilEOT, to our knowledge the first open dataset and detector for semantic end-of-turn detection in a South Indian language: 18,485 labelled turn boundaries from 116 real Tamil telephone conversations, two audio-only detectors built on Smart Turn v3's architecture, and a turn-level companion set of 4,774 whole turns with their pause structure.”
- “Data, weights, code and every negative result are public.”
来源:arXiv · cs.AI · arxiv.org