人工智能可以谱交响乐了!

    It can create digital art, write poems and now, artificial intelligence (AI) is composing music.

    Japanese researchers have developed an AI headset that creates tailor-made* music in order to improve the wearer’s mood.

    The AI analyzes the person’s brain waves and writes tunes that match their personal sensitivity and it only takes one minute to create the music using synthesized* notes.

    The AI headset was developed as a collaboration between Osaka University and Tokyo City University.

    For the first part of their work, the researchers teamed up with an institute in Belgium to record the brain waves of volunteers. These participants were each given 10 different pieces of music to listen to that ranged from J-Pop to nursery rhymes while their brain waves were recorded. The data was then used to create a personalized “emotional music model” for each person.

    The AI was fed information about the relationship between music and various emotions before it began writing music.

    According to researchers, “conventional automatic composition machines require the input of specific data about the characteristics of the music that the listener wants.”

    “I would like to turn our study into a system to uplift human spirits and help people to fulfill their potential by listening to music in accordance with their preferences,” said Masayuki Numao, a professor of information and physical sciences at Osaka University’s Institute of Scientific and Industrial Research.

    A separate team of researchers taught an AI how to compose music that sounds similar to The Beatles last year. Called “Daddy’s Car,” the tune was created by Flow Machines, a system Sony taught to make music by feeding it 13,000 samples from different genres.

    (SD-Agencies)