Decoding speech.

• Autores: Simon Makin
• Localización: Scientific American, ISSN 0036-8733, Vol. 320, Nº. 5, 2019, págs. 18-19
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • The article focuses on research into tools decoding speech-related brain activity and convert it into speech-synthesis. It states that machine learning and speech-synthesis technologies using a deep learning neural network were able to reproduce original speech 75 percent of the time. It comments that the brain activity was recorded from speech perception regions instead of speech production regions and didn't utilize complete sentences or a larger vocabulary.