Combined the artificial intelligence (AI) research teams from Google Brain and DeepMind into a single organization known as Google DeepMind. Since then, this division of the Alphabet-owned business has released a number of AI tools, and it has just unveiled its "most advanced music generation model."
The group unveiled two AI experiments and the Lyria AI music generation model in collaboration with YouTube, another Google-owned company. According to the report, the Lyria model offers users more intricate control over the style and performance of the output and can simultaneously produce "high-quality music with instrumentals and vocals."
Music, ranging from jazz to heavy metal, techno to opera, is a widely cherished medium for artistic expression. Up until now, the company claimed, artificial intelligence (AI) systems have found it particularly difficult to create captivating music because of the intricate and densely layered lyrics, melodies, rhythms, and vocals.
Google DeepMind AI experiments
Dream Track: An experiment in YouTube Shorts designed to help deepen connections between artists, creators, and fans through music creation.
Music AI tools: A set of tools we’re designing with artists, songwriters, and producers to help bolster their creative processes.
Google DeepMind said that it is trialing Lyria in Dream Track experiment wherein a limited set of creators will be able to use Dream Track for producing a unique soundtrack with the AI-generated voice and musical style of artists.
The artist list includes Alec Benjamin, Charlie Puth, Charli XCX, Demi Lovato, John Legend, Sia, T-Pain, Troye Sivan and Papoose. The company says that each participating artist has partnered to test the LLM.
How Dream Track worksDream Track will allow users to enter a topic and choose an artist from the carousel to generate a 30 second soundtrack for their Youtube Short. Using our Lyria model, Dream Track will generates the lyrics, backing track and AI-generated voice in the style of the artist selected.
Furthermore, the team will also watermark AI-generated audio with SynthID tool, which Google launched earlier this year for Google’s AI image generator Imagen. SynthID converts audio into a visual spectrogram to add a digital watermark. The tool is aimed at helping people tell when AI-generated content is being passed off as real, and/or help protect copyright.