Google has unveiled Lyria 3, its most advanced generative AI music model, integrated into the Gemini chatbot app. Users, aged 18 and over, can create 30-second tracks by inputting text prompts or uploading images. Developed by Google DeepMind, Lyria 3 supports various languages, including English, Spanish, and Japanese, and is designed for casual use, enabling personalized creations like birthday songs or pet-themed ballads. The model offers enhanced features such as automatic lyric generation and improved creative control, building on previous iterations. Google emphasizes responsible development, ensuring training materials respect copyright laws and artist agreements. Lyria 3 is also integrated into YouTube’s Dream Track, which supports original soundtracks for creators. Additionally, Google’s watermarking technology, SynthID, assists in identifying AI-generated content, and users can check if their files were created using Google AI. This launch follows a licensing deal with Universal Music Group, enhancing protections for artists in the realm of generative AI.
Source link
