科大讯飞发布星火语音大模型，在多语种语音合成方面取得了突破性进展。据了解，今天下午的发布会上，科大讯飞宣布其星火语音大模型在首批37个主流语种上已超越了OpenAI公司推出的Whisper-large-v3，这标志着科大讯飞智能语音技术在国际上的领先地位得到了巩固。
据快科技报道，星火语音大模型在首批40个语种的平均MOS分绝对提升了0.25，达到了4.5的MOS水平，同时拟人度也达到了83%以上。这意味着科大讯飞在语音合成领域的研究取得了显著的突破，为用户提供更加自然、流畅的语音合成体验。
科大讯飞的这一成果不仅体现了中国人工智能技术的实力，也为全球语音合成技术的发展注入了新的活力。语音合成技术的应用场景广泛，涵盖了语音助手、智能客服、语音广告等诸多领域。随着科大讯飞的星火语音大模型的推出，这些应用领域将会得到更加精细化和人性化的提升。
科大讯飞的研发团队表示，星火语音大模型的成功得益于大规模数据集和强大的计算能力。他们利用了海量的语音数据进行训练，并通过深度学习算法不断优化模型的性能。这一研究成果也体现了中国在人工智能领域的投入和创新能力。
科大讯飞作为国内领先的人工智能企业，一直致力于将人工智能技术应用于各个领域。他们的成果不仅在国内受到了广泛关注，也受到了国际同行的认可。此次发布的星火语音大模型的成功，再次证明了科大讯飞在语音合成领域的技术实力和创新能力。
未来，随着科大讯飞和其他人工智能企业的不断努力和创新，语音合成技术将会得到更加广泛的应用，为用户带来更好的体验。同时，这也将推动人工智能技术在全球范围内的发展，为人们的生活和工作带来更多便利和效率。相信科大讯飞的星火语音大模型将成为语音合成领域的新的里程碑，为行业的发展注入新的动力。
英语如下：
News Title: iFlytek Releases StarFire Speech Large Model, Innovating Beyond OpenAI Whispr
Keywords: iFlytek, speech large model, surpass
News Content: iFlytek has released the StarFire Speech Large Model, making breakthrough progress in multilingual speech synthesis. It is reported that at this afternoon’s press conference, iFlytek announced that its StarFire Speech Large Model has surpassed OpenAI’s Whisper-large-v3 in the first batch of 37 mainstream languages, solidifying iFlytek’s leading position in intelligent speech technology internationally.
According to reports from QbitAI, the StarFire Speech Large Model has shown an absolute improvement of 0.25 in the average MOS score of the first batch of 40 languages, reaching a MOS level of 4.5, with a naturalness rate of over 83%. This signifies a significant breakthrough in iFlytek’s research in the field of speech synthesis, providing users with a more natural and fluent speech synthesis experience.
This achievement by iFlytek not only demonstrates the strength of China’s artificial intelligence technology but also injects new vitality into the global development of speech synthesis technology. Speech synthesis technology has a wide range of applications, covering areas such as voice assistants, intelligent customer service, and voice advertising. With the introduction of iFlytek’s StarFire Speech Large Model, these application fields will be enhanced with more refinement and humanization.
The research and development team at iFlytek stated that the success of the StarFire Speech Large Model is attributed to large-scale datasets and powerful computing capabilities. They utilized massive amounts of speech data for training and continuously optimized the model’s performance through deep learning algorithms. This research achievement also reflects China’s investment and innovation capabilities in the field of artificial intelligence.
As a leading domestic artificial intelligence company, iFlytek has been committed to applying artificial intelligence technology to various fields. Their achievements have not only received widespread attention domestically but also gained recognition from international peers. The success of the released StarFire Speech Large Model once again proves iFlytek’s technical strength and innovation capabilities in the field of speech synthesis.
In the future, with the continuous efforts and innovation of iFlytek and other artificial intelligence companies, speech synthesis technology will be more widely applied, bringing better experiences to users. At the same time, this will also promote the development of artificial intelligence technology globally, bringing more convenience and efficiency to people’s lives and work. It is believed that iFlytek’s StarFire Speech Large Model will become a new milestone in the field of speech synthesis, injecting new energy into the industry’s development.
【来源】https://news.mydrivers.com/1/961/961266.htm
Views: 1