Auditory Attention Decoding when switching attention to abrupt sound

Recently, Auditory Attention Decoding (AAD), a method that uses information on the listener's auditory intention (electroencephalogram) to identify the attended speech stream among multiple speech, has been attracting attention. This method can enhance only the listener's attended speech stream and reduce the unattended speech stream. Although previous research has demonstrated high decoding accuracy in limited experiments for a single sound, for practical use, it is necessary to assume an environment that is closer to real life. Therefore, in this study, an experiment was conducted in a situation in which a person is abruptly spoken to was reproduced. Specifically, two different speeches were presented first, and after a certain period of time, a third speech was presented abruptly. The results showed a promising accuracy of approximately 60%, which is above the chance level. In addition, the attention switching resulted decreased accuracy after the switching.