Many Countries flags

How Does Real-Time Language Translation Work?

While the language barrier hasn’t been shattered, it’s definitely cracked. More technologies continue to emerge, delivering translation solutions that used to be relegated to sci-fi.

AI-powered software and applications, earbuds, computers and smart speakers allow people to speak with one another in their native languages, effectively removing the language barrier. Several companies have developed tech that can translate dozens of languages, with over 90% accuracy in some circumstances.

Williams AV launched Convey Video, a real-time translation solution for virtual collaboration.

While this technology has been around since about the 1970s — or at least the idea has been — it wasn’t until recently they’ve become more advanced.

Google’s Translate app is among the leading and most well-known. The app boasts more than 1 billion daily active users, most of which (95%) come from outside of the U.S.

Its history reflects the developments of the technology. It wasn’t always as accurate as it is today, but it continues to improve as the company sees the rising need for translation solutions.

About 50% of the content on the internet is in English. But, only 20% of the world speaks English. “If you look at areas where there is a lot of growth in internet usage, like Asian countries, most of them don’t know English at all,” said Barak Turovsky, Google Translate’s director of product, in an article. “So in that regard, breaking language barriers is an important goal for everyone – and obviously for Google.”

As the world becomes more connected through social media, email, video conferencing and myriad other technologies, it’ll become more important to break down barriers and freely communicate.

Those in higher education, government and international business stand to benefit the most from real-time language translation, whether it’s to teach the 1 million international students in the U.S., establish partnerships abroad, or enter new markets.

How Does Real-Time Language Translation Work? And What are the Benefits?

Historically, real-time translation was slow moving (at least compared to normal speech) and included a higher percentage of inaccuracies compared to today’s tech. The way software or devices translated language was by taking it one piece at a time, and then translating exactly what was said.

“Translation is typically a literal interpretation of what’s there as opposed to the meaning and the context,” said Rob Thomas, general manager of IBM Data and Watson AI, in a USA Today article.

With the inclusion of artificial intelligence and machine learning, real-time language translation has become far more advanced. Not only does it translate what’s spoken, but does so more accurately by interpreting and understanding what the speaker says.

Using Google’s Interpreter Mode on a smartphone, one can translate up to 44 languages in real-time, with accuracy above 90% for some translations. Other applications can translate better than humans can.

The level of accuracy depends on several factors, including:

  • Ambient noise. If there is a lot of background noise, it can hamper some translation tools.
  • Speech variances. The speed of talking, accents, and similar factors can play into the level of accuracy.
  • Data availability. For some translations, like English to Spanish, there is more data available, which makes the translations more reliable.

In time, language translation technology will improve, making it a leading solution for applications ranging from university classrooms and corporate meeting rooms to spontaneous collaboration.

Real-time language translation technology will continue to develop and mature as leading companies like Google, Microsoft and several other emerging companies establish products that serve those that need it most.

Real-Time Language Translation Technology Benefits a More Connected World

The technology imagined by sci-fi writers of the past inspired enough engineers and scientists to set to work, creating generations of technology that didn’t seem possible. Until it was.

Williams AV engineered Convey Video, a real-time translation solution that supports up to 27 languages and 70 dialects. Convey Video takes audio from a mixer and uses an AI engine for real-time translation captioning. The translation then overlays on video as open captions, making it easier for those in the audience to understand what a speaker says.

Convey Video is best suited for virtual presentations, like webinars or online events, distance education and corporate training. This solution strives to make virtual presentations and communication more inclusive by breaking down language barriers.

Learn more about Williams AV Convey Video solution online.