Gemini AI Equipped with Ability to Create Songs from Text Descriptions
Jakarta (ANTARA) - Google has introduced an artificial intelligence-based music creation feature into the Gemini app, allowing users to compose songs simply by writing a description.
According to a Tech Crunch report on Wednesday (18/2), the feature, still in beta, is powered by the Lyria 3 generative model developed by Google DeepMind. Through this feature, users can describe the song they wish to create, and the system will produce a track of approximately 30 seconds, complete with lyrics.
Users can also upload photos or videos, after which the system will create a song that matches the mood of the content. An illustrative song cover is also generated automatically.
According to Google, Lyria 3 represents an improvement over previous model generations, producing compositions deemed more realistic and complex. Users are also given control to adjust the style, vocals, and tempo of the song.
In addition to being available in the Gemini app, the Lyria 3 model has been extended to the YouTube platform through the Dream Track feature aimed at content creators. Whereas Dream Track was previously only available to creators in the United States, Google has now expanded its availability globally.
Google emphasised that the feature is designed to encourage original expression rather than imitate specific artists. If a user includes a musician's name in the prompt, Gemini will produce a song with a style or arrangement inspired by that artist, rather than copying existing works.
The company also stated that it has implemented filters to check the similarity of AI-created songs against previously published content.
All songs created using Lyria 3 will carry a digital SynthID label to identify AI-generated content. Google has also added a capability within Gemini to help users verify whether a music track was created with AI using this technology.
The music creation feature has been launched for Gemini users aged 18 and over globally, with support for a number of languages including English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese.
The development of AI-based music has elicited a range of responses across the industry. A number of platforms such as Spotify and YouTube have begun adopting AI technology and forging partnerships with music labels for monetisation.
On the other hand, AI model development companies face lawsuits related to the copyright of training materials. Platforms such as Deezer have also released tools to identify AI-generated music in order to curb potential misuse and streaming manipulation.
According to a Tech Crunch report on Wednesday (18/2), the feature, still in beta, is powered by the Lyria 3 generative model developed by Google DeepMind. Through this feature, users can describe the song they wish to create, and the system will produce a track of approximately 30 seconds, complete with lyrics.
Users can also upload photos or videos, after which the system will create a song that matches the mood of the content. An illustrative song cover is also generated automatically.
According to Google, Lyria 3 represents an improvement over previous model generations, producing compositions deemed more realistic and complex. Users are also given control to adjust the style, vocals, and tempo of the song.
In addition to being available in the Gemini app, the Lyria 3 model has been extended to the YouTube platform through the Dream Track feature aimed at content creators. Whereas Dream Track was previously only available to creators in the United States, Google has now expanded its availability globally.
Google emphasised that the feature is designed to encourage original expression rather than imitate specific artists. If a user includes a musician's name in the prompt, Gemini will produce a song with a style or arrangement inspired by that artist, rather than copying existing works.
The company also stated that it has implemented filters to check the similarity of AI-created songs against previously published content.
All songs created using Lyria 3 will carry a digital SynthID label to identify AI-generated content. Google has also added a capability within Gemini to help users verify whether a music track was created with AI using this technology.
The music creation feature has been launched for Gemini users aged 18 and over globally, with support for a number of languages including English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese.
The development of AI-based music has elicited a range of responses across the industry. A number of platforms such as Spotify and YouTube have begun adopting AI technology and forging partnerships with music labels for monetisation.
On the other hand, AI model development companies face lawsuits related to the copyright of training materials. Platforms such as Deezer have also released tools to identify AI-generated music in order to curb potential misuse and streaming manipulation.