The application to the silent ones

But 51-year-old father of four keeps talking, using his voice instead of some robot voice, thanks to an app developed by two Czech universities.

“seems very useful to me”, Gular told the AFP news agency, while writing down what he wants to say and then with the help of the app, the message is back in voice.

“I'm not very good at using voice prosthesis”, he said, touching the place where surgery was performed on his throat and now it's empty.

Prothesis, a small device attached to the throat, allows people to speak by hand so that there is air flow and sound.

But Gular prefers new technology sound equipment.

It was created for patients who lose their voice after laryng leaves, a procedure that people with throat cancer face.

The joint project of the University of Western Bohemia at Pilsen and Carl University in Prague has been realised along with two other private companies two years ago.

Technology uses sound recordings of patients to create a synthetic speech that can be used on mobile phones, tablets, and laptops.

The ideal is for patients to record more than 10,000 sentences so that scientists can have enough material to produce synthetic voice.

“We work with individual voices, so we need a lot of sentences”, said Yindrich Matusek, an expert on text-san change, and runs a project at Pilsen University.

“

However, there are disadvantages in this process, patients facing laryng evacuation have less energy to conduct recordings because of the treatment required by the disease.

The “is usually a matter of weeks”, said Barbora Repova, a doctor at the MotoI Hospital who works on Carl University project.

“Patients have to deal with issues such as the economic situation, their life changes very quickly when they understand the disease, and the last thing they want to do is record the voice”, she has told AFP.

To address these challenges, scientists have created another option on the app, supported by the Agency for Technology in the Czech Republic.

Working with fewer sentences, usually 3,500, but sometimes with 300, this method uses advanced statistical models, such as artificial neurons.

“

If we have more data, it's better, but you can get good quality even with less data”.

These sentences have been carefully chosen, and individual voices have been recorded several times, as sentences are pronounced differently, whether at the beginning or at the end of the corresponding word, he added.

So far, Pilsen University has recorded the voice to 10 to 15 patients, according to Matousek.

In addition to Czech language, scientists have created examples in English, Russian, and Slovak.

“Complications during recording”

Gular, who is currently unemployed because of the disease, has recorded 477 sentences for three weeks, at the time he understood the disease to the operation.

During that time, however, he has been stressed and is not satisfied with the quality of his voice.

The cancer patients with throat cancer usually suffer from voice speech, so in combination with less opportunity to speak, this affects the voice not to be neutral”, Repova said.

Meanwhile, at a studio at Pilsen University, entrepreneur Jana Huttova is recording several phrases.

The mother of three and 34-year-olds, she has lost her voice after surgery.

I have small children and I want them to hear my voice, not a robot”, Huttova said.

Relation to the Brain

Matousek believes that in the future, patients will be able to use their voice at home, using a website that will guide them for the whole process.

He even hopes that it will one day become even greater.

The final version is an extremely small device connected to the brain, respectively with nerves responsible for speaking, so patients can control the device through their” opinions, she said.

This advanced solution will take many years, Repova said.

It is real: it may not happen in a year or in 10 years, but it is possible and we are working on it”, it is expressed.