-->
Save your seat for Streaming Media NYC this May. Register Now!

What’s Next for AI Dubbing in the Media Industry?

Article Featured Image

随着最近开发的解决方案得到广泛应用,人工智能配音将极大地改变广播媒体行业. 

Numerous dubbing startups have joined the AI gold rush, capitalizing on the demand for affordable content localization. With solutions that cut production costs by 30-50%, 这些公司有潜力改变媒体和创意产业. 

目前,投资者和媒体正在努力应对评估新解决方案的挑战. However, 2024年,重点将转向新兴工具的潜在成本及其对媒体行业的影响.  

媒体本地化的主要参与者决心在AI游戏中保持领先地位. London-based VSI group, which provides media localization in over 80 languages worldwide, 聘请斯科特·罗斯担任首席技术官,以确保“VSI继续参与与配音相关的人工智能开发.”

The market for AI-powered dubbing is shaping up 

在我们的记忆中,2023年将成为人工智能语音翻译和内容本地化配音的第一批利基领导者出现的时期.  

Automated dubbing with the human touch. 这些解决方案开创了混合翻译,由人类完成人工智能配音. Startups making waves in this segment include Papercup and Deepdub (raised $20 million each in 2022). 他们的端到端翻译服务针对媒体行业,并保证主要广播公司所要求的质量.  

AI-powered DIY translation tools allow users, such as freelance content creators and small businesses, to translate their videos with AI and then make edits on their own. Such solutions, provided by Heygen, 依靠由Eleven Labs开发的自然语音合成和文本转语音软件. 

Real-time translation solutions are being developed by Zoom. 由于一些技术上的限制,我们可能要等一段时间才能取得突破.  

Voice challenges: Emotions and lip-syncing

Until recently, AI performed well for localizing “factual content”, 比如纪录片和教育节目,情绪变化较小, fewer voices, and where perfect lip sync is not crucial.

这是因为从历史上看,人工智能语音是为汽车导航辅助或机器人吸尘器设计的, and had one or two emotions. 尽管与人类的声音难以区分,但它们听起来太完美、太平淡了. 

In 2024, as technology improves, AI will be increasingly applied to movies, animation, series, and other content. 合成的声音已经变得更有情感表现力,听起来越来越像人类. 

Voice cloning has made this breakthrough possible. 现在,一个单一的人工智能模型可以支持广泛的情绪和语调. Voice cloning also helped facilitate the creation of ad-hoc voices, which was previously unavailable in AI-powered dubbing solutions. 

However, 该行业现在面临着建立法规和确保合乎道德地使用人类声音的责任. 去年发生了涉及克隆几十年前签署放弃版权的演员声音的丑闻, well before AI came into picture. 

On a brighter side, 语音克隆技术的进步有助于改善配音, for example, by correcting pronunciation.

Another important improvement is visual lip syncing, where new prototypes emerged in 2023. 它们主要是为简单的内容设计的,并且有需要人工纠正的错误. 现在,技术限制减少了,明年解决方案将进一步发展.

How will AI-powered dubbing transform the human job market?

随着越来越多的人工智能工具的出现,许多行业和职业道路都面临着根本性的转变. 以下是明年可能出现的一些与人工智能配音相关的新机会: 

AI Dubbing Manager, or proof listener, 是否会针对特定行业或内容类型微调人工智能配音系统. 这个角色可以包括听自动配音,以掌握文化的细微差别, refine voice modulation, and make corrections. 随着这个行业的发展,一些演员和口译员可能会过渡到这个行业.

Creative Directors for AI-enhanced productions will be guiding creative content developed through AI dubbing. 他们将确保技术增强而不是削弱整体的艺术和情感影响.

Actors licensing their AI-generated voices will charge fees for each usage. 更多的工具将进入市场,使个人能够通过人工智能生成自己的声音. Actors will be able to create new voices based on their own.  

Entering the last frontier for AI dubbing

今年将出现激烈的监管讨论.S. working on a new executive order and the EU preparing its AI rulebook. 这些监管差异对全球科技市场的影响还有待观察. 

随着推出新平台和服务变得更加容易,人工智能视频翻译和配音服务的数量将在2024年继续增长. Most importantly, 最后的边界——捕捉情感和调制,将人类声音与人工智能生成的声音区分开来——最终将被跨越.

[Editor's note: This is a contributed article from Dubformer. 流媒体接受供应商署名完全基于它们对我们读者的价值.]

Streaming Covers
Free
for qualified subscribers
Subscribe Now Current Issue Past Issues
Related Articles

Navigating the Future of AI for Media & Entertainment: Self-disrupt or Self-destruct?

各种规模的媒体和娱乐公司都在努力吸引和留住那些渴望找到并消费他们感兴趣的内容的观众. 由于内容丰富,可供浏览和付费的选项太多,而且转换成本很低,许多提供商无法盈利. How can media and entertainment companies respond? Many are turning to a new ally: AI.

The Video Captioning Conundrum

StreamShark的James Broberg讨论了为什么视频字幕对于视频内容的可访问性和增加观众的理解是必不可少的.

基于机器学习的视频流字幕自动对齐方法

为了确保高质量的观看体验,同时遵守地区法规,音频和字幕必须保持一致. 这可以通过利用机器学习的自动校准系统高效且经济地实现. 其结果是一种满足当今全球观众高期望的观看体验,并推动了增长.