Voice scientist Rupal Patel speaks of how voice surrogates can help people with voice disorders get an all new identity.
We see several people living with speech disorders around us. Most of these people don’t have the luxury of owning modern devices which convert text into speech. And even if they do, these machines lack the human touch. By that, we mean that the voices used by these machines are generic and common to practically everybody who uses it.
Even though it’s a male or a female talking, what we notice is that speech isn’t distinguishable based on the person. Rather, it’s more of an artificial voice which these machines give out, essentially meaning that if two people with speech disorders were to have a conversation (using computerized devices), they will sound the same.
But Rupal Patel, who is a voice scientist and a professor at the Northeastern University, has done some research in this field to find a solution for this. She has looked closely into all factors of how patients with speech disorders behave and what could be done to fix it. Her idea revolves around using a voice surrogate of the same age and gender as the target person and then using their filters, while at the same time using the voice from the target source. Then, the filter and the voice of the source can be mixed to create a voice for the patient.
The surrogate will have to dedicate a few hours of his or her time uttering out unique sentences or phrases (like shown in the picture above). Since it’s the sound that matters here, the goal is to get as many combinations of sounds in word as possible, in order to give the target a broader collection of voices. The end result is absolutely fantastic as you can see from the video below. To put it simply, all they need is a voice from the target, even as much as a vowel or a sound and the pitch of the voice, and the filter or the voice from a surrogate and the person will have their own unique voice to use.
This will help build individuality among people with speech disorder and they will sound just like normal people. Perhaps the only thing holding people back from this is the fact that these devices used to communicate among speech disordered people usually cost close to $10,000. So this might not be for everybody as of now. But given the rapid progress in the technology sector, we might be able to see such devices being available for much lesser.
The team at VocaliD is looking for contributors all over the world and you can contribute through their site at vocalid.org. It is currently functional only for Windows, but the folks hope to get Android and iOS apps up and running soon to get more contributors on board with this excellent project.