RESUMO
Total removal of the larynx may be required to treat laryngeal cancer: speech is lost. This article shows that it may be possible to restore speech by sensing movement of the remaining speech articulators and use machine learning algorithms to derive a transformation to convert this sensor data into an acoustic signal. The resulting "silent speech," which may be delivered in real time, is intelligible and sounds natural. The identity of the speaker is recognisable. The sensing technique involves attaching small, unobtrusive magnets to the lips and tongue and monitoring changes in the magnetic field induced by their movement.
Assuntos
Acústica/instrumentação , Laringectomia , Laringe Artificial , Lábio/fisiologia , Aprendizado de Máquina , Magnetismo/instrumentação , Imãs , Acústica da Fala , Língua/fisiologia , Transdutores , Qualidade da Voz , Fenômenos Biomecânicos , Humanos , Campos Magnéticos , Desenho de Prótese , Recuperação de Função Fisiológica , Processamento de Sinais Assistido por Computador , Espectrografia do Som , Inteligibilidade da Fala , Fatores de TempoRESUMO
By speech articulator movement and training a transformation to audio we can restore the power of speech to someone who has lost their larynx. We sense changes in magnetic field caused by movements of small magnets attached to the lips and tongue. The sensor transformation uses recurrent neural networks.