INDUSTRY REPORT

摘要

Columbia University neuroen-gineers have scored a scientific first with their system that translates brain waves into recognizable speech patterns with unprecedented clarity. Combining the technologies of artificial intelligence with speech synthesizers, the discovery could lead to other ways for computers to interact with the human brain. An obvious advantage would be to allow individuals who have lost the ability to speak, to interact with the outside world. Nima Mesgarani, PhD, and his team at Columbia University's Mortimer B. Zuckerman Mind Brain Behavior Institute, employed a vocoder-a computer algorithm, and the same technology used by Amazon Echo and Apple's Siri-to replicate sounds based on measured brain patterns of epilepsy patients to spoken simple word stimuli. The vocoder's sounds were analyzed and cleaned up by neural networks.