ESPnet-ST
收藏arXiv2025-09-30 收录
下载链接:
https://espnet.github.io/espnet/
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是一个专注于端到端语音到文本翻译的工具包,它采用了改进的跨注意力块。此外,本文中的模型是以预先训练的自动语音识别模型为基础进行初始化的。该数据集的任务是语音到文本神经机器翻译(S2T NMT)。
This dataset is a toolkit focused on end-to-end speech-to-text translation, which adopts improved cross-attention blocks. Additionally, the models in this paper are initialized based on pre-trained automatic speech recognition models. The task of this dataset is speech-to-text neural machine translation (S2T NMT).
提供机构:
National Institute of Advanced Industrial Science and Technology (AIST)



