Ai-Media consistently deliver accurate video captioning in a range of languages

The Company

Ai-Media provides valuable captioning and transcriptions across the globe and required a partner to provide accurate speech recognition technology to deliver the best quality captions for live broadcast media as well as recorded video content.

The Challenge

Video captions are essential to people who are deaf or hard of hearing, however, they are now also serving wider demand. Today, people are consuming more video content on their mobile devices in noisy environments, and they rely on captions so they don’t miss a single word. Content creators recognize that all audiences tune in more often and for longer periods when captioning is available, and the video content that has the farthest reach and impact includes captions. The BBC’s research found that 80% of its survey respondents who used closed captions had no hearing loss and 70% said closed captions improved their comprehension.

The Solution

The partnership was developed with Speechmatics because of its excellence in language coverage for automatic speech recognition (ASR). Speechmatics has a Global English offering, which is one package for all accents and dialects of English, including American, British, Australian and others. Speechmatics’ ASR technology is also particularly strong with Spanish, French, Korean, Mandarin, Japanese, Malay and Nordic languages, which are all very important to Ai-Media’s customers for captioning, subtitles and translations. Ai-Media’s customers span the globe and rely on Ai-Media for the company’s vast language offering.

Ai-Media has integrated Speechmatics’ any-context speech recognition engine into their workflow to consistently deliver accurate video captioning in a range of languages. The service offers top-quality captions that are easy to read and convey depth-of-meaning, so all audiences receive the best experience.

The Benefits

Ai-Media’s customers reap benefits when the transcription or captioning process starts with Speechmatics. The pre-recorded videos initially pass through Speechmatics’ any-context speech recognition engine, which leverages machine learning. Then, Ai-Media takes that output and creates the transcription and captions. Ai-Media’s proprietary software understands language and phrasing and enables quality control, editing and formatting for a polished final product. Customers quickly get exemplary transcriptions and captions accompanying their video content that are clear, easy and accurate for audiences.

The Results

Accuracy is one key measure of video captions and Speechmatics’ ASR engine contributes to that piece of the puzzle in Ai-Media’s offering. But accuracy alone is not enough to produce truly great captions. How captions are displayed has as much impact. It’s critical that the flow of the captions feels natural and provides the same emotional queues that hearing audiences receive. Ai-Media has developed technology that understands how phrases need to be broken up into lines so that the meaning can be ascertained quickly and the dialog flows naturally at the correct time. ASR gives you the words, and Ai-Media’s processing of those words results in an artfully delivered experience.