University of Columbia University has developed a system of speech signals that are intelligible, recognizable speech. By clarifying clarity. The breakthrough reported in the journal Scientific Reportscouldn’t be able to speak.
“If you’re a little boy, do you?” With today’s study, we’ve got We’ve shown this Nima Mesgarani, Principal Investigator for the Mortimer B. Zuckerman Mind Brain Behavior Institute at Columbia University.
Early efforts to decode He has analyzed the sound patterns that have been analyzed.
But it’s not a problem.
“This is the same technology used by Amazon to support your questions,” Dr. Mesgarani said.
“What patterns and brain activity.” These neural patterns trained the vocoder. ”
Next, Record Digits between 0 and 9, while he was listening through the vocoder.
It has been shown that there has been no need for any further information. The end result was a robotic-sounding voice reciting sequence of numbers.
He heard what he heard.
“Attempting to find out about Mesgarani said.
“For comparing to previous, spectrogram-based attempts. ”
It is a question of a person who speaks or imagines speaking.
Ultimately, it’s not a problem.
Hassan Akbari et al. 2019. Towards reconstructing cortex. Scientific Reports 9, article number: 874; doi: 10.1038 / s41598-018-37359-z