Voice Navigation
收藏arXiv2025-09-30 收录
下载链接：
https://1drv.ms/u/s!AmYoFTLtyCO_k49SFLCJOneZBYCWCA?e=2GgEcS
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个大规模的中文语音填槽数据集，包含了超过83万个地点名称和生成的查询样本。数据集不仅包括通过文本转语音（TTS）生成的数据，也包含了真人朗读的数据。此外，测试集中有一半的槽位是在训练集中没有出现过的词汇。该数据集的规模包括超过83万个样本用于训练，以及12,000个样本用于测试，其任务专注于语音填槽。
This is a large-scale Chinese spoken slot filling dataset containing over 830,000 location names and generated query samples. It includes both data generated via Text-to-Speech (TTS) and human-recorded speech data. Additionally, half of the slots in the test set consist of vocabulary that does not appear in the training set. In terms of scale, the dataset comprises over 830,000 training samples and 12,000 test samples, with its task focusing on spoken slot filling.
