Micah Fowler, left, and Cedric Yarbrough participate in the "Speechless" panel during the Disney/ABC Television Critics Association summer press tour in Beverly Hills, California, August 4, 2016.
Ryan. Alex. Siri. These are names of robotic voices that are often used on cellphones or tablets. They have also become voices for people with speech problems who use technology to talk.

Now, some speech scientists are developing custom voices to reflect the diversity of the people who use them. To do it, they are tapping into a network of volunteers who have donated their voices to share with people who cannot speak.

17,000 Volunteers Donated To Voicebank

The effort to build an international "Human Voicebank" has attracted more than 17,000 volunteers from 110 countries. One is Priyanka Pandya, a 16-year-old from Columbia, Maryland. She plans to spend her winter vacation recording a string of sentences into her laptop.

Her voice could be used to help one of more than 2 million Americans who have severe speech problems and need help to communicate. It's a problem captured in the television series "Speechless." The main character is a teenager with cerebral palsy, which affects movement and muscle use. He relies on technology and an assistant to act as his "voice" at school.

"Everyone has a voice," said Rupal Patel, founder of VocaliD, the Belmont, Massachusetts, company that launched the voicebank. "Even people who are speechless have sounds that are unique to them."

Her company designs custom voices by recording the sounds of the user. Then those sounds are blended with six to 10 hours of recordings from a voice volunteer, matched to the user.

Researchers Already Using Speech Technology

The company is developing voices now for its first 100 customers. However, researchers have been perfecting the technology for years.

Tim Bunnell is head of the Speech Research Laboratory at the Alfred I. DuPont Hospital for Children in Wilmington, Delaware. He has created voices for more than 1,000 people with diseases such as amyotrophic lateral sclerosis, or ALS. Those people were usually able to record their own voices before they lost the ability to talk.

The scientists in Bunnell's lab study the user's sounds for qualities such as pitch. Then they map out the recordings from a voice volunteer. They combine the voices by changing the volunteer voice to reflect the qualities of the user's voice.

Bunnell said they work to match the "vowel quality" because the "color" of someone's voice is mainly heard in the vowels.

Patel is also a speech technology professor on leave from Northeastern University. She worked with Bunnell as a researcher. When she decided to bring the technology out of the lab, she developed her own method for blending voices and turned to the online world for help. She wanted to have enough voices to reflect the diversity of customers, which could include a man from Alabama or a little girl from Ireland.

Volunteers Participate For Various Reasons

Since she announced the voicebank two years ago, the response has been overwhelming, she said, with volunteers ages 6 to 91 logging on all over the world. People help for many reasons. Some are practicing English or performing community service. Some have throat cancer or diseases such as ALS and could eventually receive voice help themselves.

Pandya began recording her voice last summer, using headphones and speaking into her computer as she read from her screen. Then she asked family, friends and others to do the same.

Patel said she is continuing to simplify the process to build voices efficiently. The $1,249 cost puts it out of reach for many people. For volunteers, it can take six to 10 hours to record the 3,500 phrases required to complete a voice sample. Only about one out of 10 volunteers makes it all the way through.

Pilot Program Participant Satisfied

Some of the first customers say they are happy with the results.

John Gregoire, of Windham, Maine, was one of the first to receive a customized voice last December from VocaliD. It was eight years after he was diagnosed with ALS. More than six years had passed since his voice had become unclear to everyone except his wife and youngest son.

His wife, Linda, said she got used to hearing him speak with "Ryan," the voice in his computer. When they heard his new voice, they reacted together instantly. Linda said, "It was him."

The couple is raising money to build a studio so people in their community can record their voices.

"Having a distinctive voice is like getting something back that was stolen," John said.

]]>