With the help of text cues, Google’s new AI tool is now able to compose music of any genre. It can even translate melodies from one instrument to another when they are hummed or whistled. According to Google Research, the recently released method, known as MusicLM, is a text-to-music creation system. The text is examined by the AI technology to determine how lengthy and intricate the composition is.
“We introduce MusicLM, a model generating high-fidelity music from text descriptions such as a calming violin melody backed by a distorted guitar riff’,” the research paper mentions. It further added, “We demonstrate that MusicLM can be conditioned on both text and a melody in that it can transform whistled and hummed melodies according to the style described in a text caption,”.
According to the research article, the new artificial intelligence has been trained on a dataset of 280,000 hours of music in order to discover details like mood, melody, and instrumentation and learn how to generate songs that make sense from word descriptions. With it, you can do more than just make song snippets. Google researchers proved that the algorithm can build on modern songs, whether they are hummed, sung, whistled, or played on an instrument.
Additionally, the study demonstrates that Google’s new AI tool is capable of taking a list of written instructions, such as “time to meditate,” “time to wake up,” “time to run,” and “time to give 100%,” and converting them into a melodic “story” or narrative that can endure for several minutes. Furthermore, Google’s new AI tool might be guided by a picture and caption combination or it might generate music that is “played” by a specific kind of instrument in a specific game. It should be mentioned that Google is not the first company to do this.
Initiatives like Google’s AudioLM, OpenAI’s Jukebox, or Riffusion—an AI that can make music by simply looking at it—have all made an attempt, according to TechCrunch. Due to technical limitations and a lack of training data, none have been able to create songs that are incredibly difficult to create in high-fidelity. Researchers believe that MusicLM may be the first to be able to do this.
