今日下午，人工智能领军企业科大讯飞召开发布会，宣布其星火语音大模型取得重大突破。该模型已成功支持首批37个主流语种，整体性能超越了OpenAI的Whisper-large-v3，标志着我国在智能语音技术领域继续保持国际领先地位。
据悉，星火语音大模型集成了多语种语音合成技术，不仅在语言数量上有所增加，更在质量上实现了飞跃。据发布会上公布的数据，该模型在首批40个语种的平均MOS分（音质评价指标）提升了0.25，达到了4.5的高分，同时拟人度也超过了83%。
此次星火语音大模型的突破，是我国人工智能领域的一大里程碑。它不仅体现了我国在人工智能技术上的实力，也为全球多语种交流提供了强有力的支持。在全球化日益深入的今天，这一成果显得尤为重要。
英文标题：Keduo Xunfei’s Spark Voice Model Surpasses OpenAI in 37 Languages
关键词：Keduo Xunfei, Spark Voice, Multilingual Breakthrough
英文新闻内容：
This afternoon, leading AI company Keduo Xunfei held a press conference to announce a significant breakthrough in its Spark Voice Model. The model has successfully supported the first batch of 37 mainstream languages, outperforming OpenAI’s Whisper-large-v3, maintaining Keduo Xunfei’s international leadership in intelligent voice technology.
It is understood that the Spark Voice Model integrates multilingual speech synthesis technology, not only increasing the number of languages but also achieving a leap in quality. According to the data announced at the press conference, the model has improved the average MOS score (a measure of sound quality) by 0.25 to a high score of 4.5 in the first batch of 40 languages, while the human-like degree has exceeded 83%.
This breakthrough of the Spark Voice Model represents a major milestone in China’s AI field. It not only reflects China’s strength in AI technology but also provides strong support for global multilingual communication. In today’s increasingly globalized world, this achievement is particularly important.
【来源】https://news.mydrivers.com/1/961/961266.htm
Views: 1