Participants 2022

Our consensus is that AI would be a good assistant in the music-making process, in ways of increasing the efficiency and creativeness of human musicians. In AISC2022, we mainly focused on performing the entire vocal-related process through three AI assistants. Furthermore, we wanted to experience the actual workflow of working with them. To achieve AI vocal-making process, three Deep Neural Network models were used to make lyrics, melodies, and singing voices.
The vocal is the most critical and trickest part of modern pop music production. Lyrics, melodies, and singing voices with the proper mood are needed to complete the songs. The process almost decides the song's identity, sometimes even more important than instrument tracks. As we have to write lyrics and melodies and need thoroughly to find appropriate vocalists, the vocal-related process is time-consuming in general. Hence, we tried to seek a way to achieve the process with the help of AI assistants. Our objectives were clear. Firstly, we wanted to achieve the vocal-related process in a time-efficient way by using Deep Neural Network models. Secondly, we wanted to discover creative results from the models beyond human musicians' thinking, such as Happy Accidents, which means unexpected creativeness. Thirdly, we tried to find whether the AI outputs could be a new source of our inspiration.