I open my mouth to speak, but my lips don’t close when they need to, and my tongue doesn’t move where I tell it to; strangers don’t understand me.
This is an everyday occurrence for people with motor speech impairments which may be associated with cerebral palsy, stroke, Down’s Syndrome, Parkinson’s Disease and ALS. Speakers try to communicate with new acquaintances but fail to convey their message. Even smart home devices can’t understand them. A simple command like “play my favorite song” goes unanswered. This is a reality that many people with speech disabilities contend with daily, but Voiceitt is here to change that.
Voiceitt is a voice technology startup and social enterprise based in Israel. Fifteen years ago, Voiceitt Co-Founder Danny Weissberg’s savta (grandmother) had a stroke, and he distinctly remembers the pain and sadness he felt because he could no longer communicate with her. Savta would open her mouth to speak, but the family could not understand her. The only person who knew exactly what Savta was saying was the nurse. That was Danny’s “eureka” moment: if the nurse could understand Savta and many others like her, then AI (artificial intelligence) could be trained to do the same. That’s when the idea for Voiceitt was born and developed in Ramat Gan.
Voiceitt uses AI as a means of understanding non-standard speech patterns and converting them into easily understood speech and text. Currently, Voiceitt has a free iOS app with two main functions: interpersonal communication and smart home control with Alexa integration. The unique integration with Alexa allows users to communicate directly with Alexa through the Voiceitt app, making the interactions seamless and efficient. The app requires the user to train each phrase and use them to communicate with others and their smart home technology.
A benefit of the app for users who find speaking an effort is that the user can train short words or phrases to trigger the app to output longer sentences. For example, if the user wants to order a hamburger made a certain way, all he needs to do is train the word “hamburger” and program the app to “say” the extended order out loud.
Alex Lytwin, a Voiceitt user, described the experience of living with a disability and how Voiceitt helped him: “Having cerebral palsy and a speech impairment can cause many struggles on a daily basis. However, Voiceitt allows me to speak with ease and lets the world see my entire life’s potential.” For some, the current app doesn’t quite meet their needs.
Barbara, who was diagnosed with ALS, a disease causing progressive deterioration of the muscles, wrote, “I’m seeking deeper communication with my partner and friends.” Barbara will be among the first users to test the next generation of Voiceitt innovation.
Voiceitt has been collecting non-standard speech samples in a proprietary database called Ensemble. The database allows the automatic speech recognition technology to learn the different ways people can pronounce words and learn to recognize non-standard speech. As the database increases and expands with more audio data, Voiceitt’s technology advances. Voiceitt is inviting people with non-standard speech (clinically called “dysarthria”) to join Ensemble. Those who join will be a part of the Ensemble community who have done their part in enhancing speech recognition for all while removing barriers to communication. Ensemble members, by recording their voices, will have the chance to be among the first to experience the newest technology by joining a closed cohort of beta users.
These beta users are currently testing the next generation of speech recognition technology, offering select beta testers and those who have completed 1,000 recordings in Ensemble a unique spontaneous speech recognition web-based tool that will work on any device that can open a web browser. The technology will allow speakers to communicate freely on video calls with live transcriptions, transcribing emails or ordering a pizza.
Voiceitt was born out of a seemingly impossible dream to create a communication solution for millions around the world with non-standard speech who aren’t being served by existing voice products by developing technology to support their unique speech patterns. Voiceitt’s partners include Yoocan, an Israeli-based empowerment and community resource for people with disabilities and their families, and AMIT Technion, which invests in exceptional biomedical innovations conceived by Technion students, faculty and alumni (co-founders Danny Weissberg and Stas Tiomkin are both alumni). Voiceitt has also received investment from Amazon, Microsoft, and AARP and is working with state governments, such as the Department of Developmental and Intellectual Disabilities in Tennessee, to bring the technology to people with speech and motor impairments so they can contribute to society and enter the workforce.
Become part of this effort to bring accessible speech recognition technology to all, regardless of speech ability. Voiceitt invites people with dysarthric speech to join Ensemble and contribute to the next generation of continuous speech recognition. Additionally, Voiceitt seeks investors to join its efforts in building sophisticated technologies that remove barriers to communication for all. Voiceitt has often worked with college interns seeking to be a part of the voice equity revolution and to join an impactful Israel-based startup.
For those who have experienced a breakdown in communication with a loved one or who have experienced the frustration of not being understood, Voiceitt is here to help them be listened to and understood. Join us.
Dr. Rachel Levy is a practicing speech-language pathologist and customer success manager at Voiceitt. Connect with her at [email protected].
