Loss of speech and the ability to speak can be a result of many mental or physical problems in the body. A stroke, brain damage, tumour in ...
Loss of speech and the ability to speak can be a result of many mental or physical problems in the body. A stroke, brain damage, tumour in the brain, neuron damage and many others can cause this speech impediment. It is all in the brain, they do want to say something, they do want to communicate like any of us. The technology of using brain waves to move objects on command is already in existence. What if thoughts could be read out loud?
Neuro-engineers from the University of Columbia, New York were working on such technology to make a way for the ones with speech difficulties to speak their mind. They have developed a new technology with the use of Artificial Intelligence in converting thoughts to speech.
The neural networks are studied and they are converted to speech using brain signals. The outcome of the theoretical part in practicality did good 80% of the time. The tech was able to produce understandable and relevant sounds according to the speaker's thoughts.
But first, Artificial Intelligence has to study the patterns of speech and reading of the speaker to create an algorithm for translation. After it keeps track of the brain waves and the outcomes, it combines and coordinates them. Later it translates them to intelligible sentences according to the patterns and tries to mimic them out loud.
The paper's senior author Dr Nima Mesgarani says, "We've shown that, with the right technology, these people's thoughts could be decoded and understood by any listener". The trials to make the tech work has faced many failures in its initial ones.
It is after they have adopted the use of an 'algorithm' that this was successful. The computer algorithm they used here is called a 'vocoder'. It is a kind of an algorithm that gets more 'trained' and improves for every record that it embeds in its system.
"This is the same tech used by Amazon echo and Apple Siri to give verbal responses to our questions...We asked epilepsy patients already undergoing brain surgery to listen to sentences spoken by different people, while we measured patterns of brain activity. These train the vocoder".
Neuro-engineers from the University of Columbia, New York were working on such technology to make a way for the ones with speech difficulties to speak their mind. They have developed a new technology with the use of Artificial Intelligence in converting thoughts to speech.
The neural networks are studied and they are converted to speech using brain signals. The outcome of the theoretical part in practicality did good 80% of the time. The tech was able to produce understandable and relevant sounds according to the speaker's thoughts.
But first, Artificial Intelligence has to study the patterns of speech and reading of the speaker to create an algorithm for translation. After it keeps track of the brain waves and the outcomes, it combines and coordinates them. Later it translates them to intelligible sentences according to the patterns and tries to mimic them out loud.
The paper's senior author Dr Nima Mesgarani says, "We've shown that, with the right technology, these people's thoughts could be decoded and understood by any listener". The trials to make the tech work has faced many failures in its initial ones.
It is after they have adopted the use of an 'algorithm' that this was successful. The computer algorithm they used here is called a 'vocoder'. It is a kind of an algorithm that gets more 'trained' and improves for every record that it embeds in its system.
"This is the same tech used by Amazon echo and Apple Siri to give verbal responses to our questions...We asked epilepsy patients already undergoing brain surgery to listen to sentences spoken by different people, while we measured patterns of brain activity. These train the vocoder".
COMMENTS