Photo by Tima Miroshnichenko from Pexels

Playing It By Ear – Sanas’ real-time accent translation

Perhaps you’ve found yourself in this uncomfortable place a time or two. I know I have.

It starts when something goes wrong with your computer, something inexplicable (at least to you), and everything on your standard troubleshooting list fails.

Fortunately, you have saved the phone number for your Internet provider’s customer service department. Unfortunately, after waiting 45 minutes on hold, you realize that the person who finally answers the call is speaking with a South Asian accent that is almost unintelligible. The call turns out to be a long, frustrating slog, and by the time you hang up, you’re probably thinking: “Why would a company hire someone for customer service who can’t communicate with the customers?”

Later, however, some empathy may creep in for the person on the other end of that call. What must it be like to try and help someone when virtually everything you say is followed by “What?” or “Excuse me?” or “Say that again? — or maybe by a more direct and R-rated response.

This was the direction from which Maxim Serebryakov, Andrei Perez Soderi, and Shawn Zhang waded into this problem. When the trio was attending Stanford, one of their friends told them he had lost his job working at a call center because of his accent. As foreign students themselves, his three classmates not only sympathized but saw an opportunity.

Serebryakov, Soderi, and Zhang aren’t at Stanford anymore, but they took their friend’s situation with them when they left. Eventually, they decided to create the software for a startup and accompanying app that would confront the accent issue directly.

In an article for CNN last month, Catherine Shouchiet wrote: “Now their company, Sanas, is testing out artificial intelligence-powered software that aims to eliminate miscommunication by changing people’s accents in real-time. A call center worker in the Philippines, for example, could speak normally into the microphone and end up sounding more like someone from Kansas to a customer at the other end.

“Call centers, the startup’s founders say, are only the beginning. The company’s Website touts its plans as “‘Speech, Reimagined.'”

On that Website, they explain: “Sanas is a company built on one goal: helping the world understand and be understood. We celebrate the diversity and uniqueness of accents and recognize that mismatches can get in the way of comprehension.

“We created Sanas after our own struggles with our own accents (Chinese, Spanish and Russian) and realized we could help everyone understand each other better, unlocking the true potential of global communication.”

Coupled with many other new devices designed to break down communication barriers, this is a frontal assault on a glitch that has existed since the dawn of man. According to Ethnologue, an extensive catalog of the world’s languages, humans now converse in 6,909 distinct tongues, cloaked in almost that many accents.

This is a particular problem in the United States, which is quite possibly the most diverse nation on the globe. Although most American secondary schools — and many colleges — require the study of at least one other language before graduation, studies have shown that the number of Americans who display actual fluency in a second language is generally set at just over 25 percent. Of those, 55 percent speak English and Spanish.

Of course, learning another language is time-consuming, and most Americans have no real incentive to become multi-lingual. After all, most of the world plays by our linguistic rules, English is the closest to a universal language.

During a Congressional debate on multiculturism, a congressman once declared: “If English was good enough for Jesus Christ, it’s good enough for me.”

Like all languages, however, English is prone to regional dialects and accents, both in the U.K. and the U.S.A. conversation between someone from Brooklyn and someone from Alabama could run into difficulties, for instance, as would an attempt at Scottish-Irish dialogue.

Even Alexa gets confused sometimes.

“We quickly found out, when Alexa was turning the lights on in random places in the house and making them pink, that Alexa does not understand my dad’s accent at all,” Perez told CNN.

Also, according to the article: “Currently, Sanas‘ algorithm can convert English to and from Australian, British, Filipino, Indan and Spanish accents. They can add a new accent to the system by training a neural network with an audio recording from professional actors and other data”

Don’t be surprised, then, to discover that George Clooney or Brad Pitt are helping you to fix your computer.

Of all the benefits that online technology can bring to the world, instant language translation might be close to the top of the list. Language differences have long been a barrier to international commerce and culture. Still, incremental steps such as the one taken by Sanas are edging us ever closer to the goal of universal communication.

This hope is widespread, for Sanas received $5.5 million in seed money before testing began.

Several call centers are currently trying out the app, which has thus far been marred only by a slightly robotic tone in some instances. Meanwhile, the creators of Sanas are exploring more linguistic possibilities.

Only 6, 091 languages to go.

Darrell Laurant
Founder at Snowflakes in a Blizzard | + posts

Darrell Laurant is a veteran journalist who previously worked at the News & Advance (Lynchburg). He published over 7,000 pieces in three decades. Darrell has covered papal visits, the Olympics, American sports, and political issues in Virginia. He has also written a variety of books, including "Inspiration Street: Two City Blocks that Helped Change America."

Share on social media
Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments
0
Would love your thoughts, please comment.x
()
x