据 IT 之家 9 月 25 日报道，Spotify 正在测试一项新的人工智能功能，名为“语音翻译（Voice Translation）”。该功能可以克隆播客主播的声音，并用该声音使用另一种语言读出播客。该功能由 Spotify 在 OpenAI 的自动语音识别 (ASR) 系统 Whisper 的帮助下开发，使用了语音转文本生成 AI 模型来翻译音频文件，并使用语音复制模型来匹配原始说话者的风格。目前，首批参与该功能的主持人包括戴克斯・夏普德、莫妮卡・帕德曼、莱克斯・弗里德曼、比尔・西蒙斯和史蒂文・巴特利特。Spotify 表示，未来几天和几周内将推出更多剧集，还将提供法语和德语翻译。
新闻翻译：
According to IT 之家 on September 25th, Spotify is testing a new AI feature called \”Voice Translation.\” This feature can clone the voice of a host and use it to read out the broadcast in another language. This feature was developed with the help of OpenAI’s automatic speech recognition (ASR) system Whisper and uses an AI model for speech-to-text translation and a speech synthesis model to match the style of the original speaker. Currently, the first hosts to participate in this feature include 克斯・夏普德，莫妮卡・帕德曼，莱克斯・弗里德曼，比尔・西蒙斯和史蒂文・巴特利特。Spotify said that it will release more episodes in the coming days and weeks and will also provide French and German translations.
【来源】https://www.ithome.com/0/721/612.htm
Views: 1