Vlastimil Gula's life had not been vaccinated a year ago: the small surgeon in the flowing veins showed a throat cancer and lost his larynx.
But a 51-year-old 51-year-old father is still jokingly using his voice, a robot roundabout, thanks to an innovative application developed by both Czech universities.
"It seems to me very useful," said Gular, AFP, in the voice of the application, using the mobile phone, in his voice.
"I'm not very good through the voice prosthesis," he added in the hole in a large slippery coin.
This small silicon-set device in the throat speaks to people by pressing their finger to adjust their air through the projection and adjust the sound.
But Gular prefers the most advanced voice technology.
The laryngectomy or laryngeal removal was developed for patients with a typical procedure for the throaty cancer phase.
At the West Bohemian University in Pilsen, Charles University in Prague and two private companies – CertiCon and SpeechTech have been working for two years now.
Technology uses recordings of the patient's voice to create a synthetic speech that can be played on mobile phones, tablets or portable computers.
Ideally, patients need more than 10,000 sentences to supply the necessary material for scientists to produce their synthetic voice.
"We're editing individual sounds of speech, so we need a lot of sentences," said Jindrich Matous, an expert on text speech synthesis, speech modeling and acoustics at Pilsen's university project.
& # 39; Week subject & # 39;
But there are disadvantages: patients often have little or no energy before recording laryngectomies after a diagnosis that requires rapid treatment.
"It's usually a week," said Doctor Barbora Repova Motol University, working at the Charles University project.
"Patients are confronted with issues such as the economic situation, their lives are weakening and finally they want to make the recording," said AFP.
To solve these difficulties, scientists improved the method of application, which is supported by the Czech Technology Agency.
Lesser used phrases – 3,500 but less than 300 – This method uses advanced statistical models, such as artificial neural networks.
"The word model uses some parameters to create a synthesized language," Matousek said.
"Having more data is still better, but you can get decent quality with less than one voice data".
The phrases are selected carefully and individual sounds must be recorded several times, for example, in a different language, next to different sounds or at the beginning and at the end of a word or phrase, he added.
So far, Pilsen University has registered 10 and 15 patients according to Matousek.
In addition to Czech, Pilsen's scientists have also produced sample samples synthesized in English, Russian and Slovak.
Baby Dinosaurs & # 39;
Gular – a tapestry that lost his job due to his handicap has earned 477 sentences during his diagnosis and operation for three weeks.
But he emphasized and less than happy his quality of voice.
"Cross-cancer patients often suffer a dysphonia (hoarseness) before the surgery, so with the limited expression of the speech, the sound is not natural," said Repova.
In a study at Pilsen University, businessman Jana Huttova records extra words.
A three-year-old 34-year-old mother runs her risk of losing her sore throat to her parathyroid glands.
"The Chechenians have always preferred a Kalashnikov dagger," he says before reading the text.
"I have little children and I want to hear my voice, not a robot," said Huttova.
He then moved on to his next sentence: "The baby of Tiranosaurus was attacked by dinosaurs."
Tied in the brain
Matosa believes that in the future, patients may use the application to use their own guided web site during the process of recording their voices.
One day, he expects even more.
"The final view is a miniature device connected to the brain, with regard to speech-related nerves, which can control the device with its thoughts," he said.
This advanced advance is very far away, said Repova.
"But look at coal implants – 40 years ago, we had no idea how we would develop it, how it was deployed at the end of the day," he said about internal deafness in dealing with severe deafness.
"A happy mother would be a device established in her throat to talk to a patient's voice," said AFP.
"It's realistic: it might not be in a year or ten years, but it's realistic and we're on the way."