Furthering its mission to help make content more accessible and fully understandable to a wider range of people, ENCO (Booth N2524) will take the wraps of its new enTranslate automated translation and captioning system at the 2019 NAB Show. Combining the powerful speech-to-text engine from ENCO’s patented, market-leading enCaption automated captioning solution with advanced translation algorithms, enTranslate provides near-real-time translation of live or pre-recorded content for alternative-language closed captioning, subtitling, and more.
enTranslate offers broadcasters and live presenters an easy and affordable solution to automatically translate television programming, board meetings, legislative sessions, lectures, sermons, and other content. TV broadcasters can engage wider audiences by offering additional closed caption options – such as CC2 or CC3 – in multiple languages, while government institutions, universities, houses of worship and corporations can embed translated captions in recorded content for subsequent viewing or display live, open-captioned subtitles on local displays for in-person attendees.
“Just like enCaption makes the audio element of video content accessible to hearing-impaired audiences, enTranslate makes it understandable to viewers who aren’t native speakers of the material’s original language,” says Ken Frommert, President of ENCO. “enTranslate delivers fast, high-quality translations at a fraction of the cost of traditional translation services, making live translation practical and affordable for both broadcast and non-broadcast applications. And because it’s available 24/7/365, users can translate content such as breaking news alerts or ad hoc presentations at a moment’s notice, without having to wait for a human translator to arrive.”
enTranslate builds on the highly-accurate, low-latency speech recognition core first implemented in ENCO’s award-winning enCaption closed and open captioning system. Leveraging machine learning technology and a deep neural network approach, the speech-to-text engine interprets incoming live or file-based audio in near real-time with continually improving exactitude.
enTranslate then feeds the resulting text to its advanced translation process, which supports 46 languages including English, Spanish, French and more. Blending artificial intelligence with sophisticated linguistics modeling, enTranslate uses a Neural Machine Translation (NMT) methodology to provide high-quality translations based on the context surrounding the current words and phrases.
Deployable on-premises or in the cloud, enTranslate’s flexible architecture supports a wide range of live baseband and network-based audio and video inputs including analog, AES, HDMI, SDI, AoIP, NDI, and MADI. Translated results can be output in standard caption file formats; embedded as closed captions using an external or optional integrated encoder; or keyed as open captions over an SDI, HDMI or NDI video output.
enTranslate offers both live and offline translation. For file-based applications, audio or video clips can be easily ingested into the system and translated into any supported language, enabling users to quickly and affordably process large libraries of previously recorded content.
enTranslate is expected to be commercially released in Q2, and like all ENCO solutions, will be backed by the company’s renowned, world-class, 24/7 technical support services.