In today's music creation, the application of AI technology has attracted more and more attention. However, many start-ups face huge challenges in this field due to copyright issues and reasonable compensation for artists.IndiaofStartups Beatoven.ai appears to have found an ethical and responsible way to address these challenges.
Mansoor Rahimat Khan, co-founder and CEO of Beatoven.ai, is a professional sitar player who comes from a family with a seven-generation musical tradition. He developed a keen interest in music technology while studying at the Indian Institute of Technology Bombay. He said that although there were not many opportunities in India, he wanted to combine his passion for music with technology.
Founded in 2021, Beatoven.ai has quickly attracted the attention of investors, completing a $1.3 million pre-Series A roundFinancing, the total financing amount reached US$2.42 million. Khan and his partner Siddharth Bhardwaj found that independent game developers, institutions and production companies have the greatest demand for soundtrack generation. Their goal is to provide users with a simple interface to generate soundtracks by selecting music type, mood and duration.
Despite the initial data collection challenges, Khan and his team have successfully accumulated more than 100,000 proprietary data samples by establishing partnerships with multiple artists. With the introduction of the Diffusion model, the quality of Beatoven.ai's music generation has been significantly improved. The company uses different models to complete different tasks, including using OpenAI's ChatGPT API and the CLAP model for video generation.
In addition, Beatoven.ai recently launched a feature called “Augment” that is similar to a sound wave generation model, allowing users to create remixes based on existing music. Khan showed how to turn a simple sitar melody into a hip-hop-style remix, fully demonstrating the flexibility of creation.
Currently, Beatoven.ai is also testing a video-to-audio model, where users can upload videos and the model will generate corresponding music based on the content. This feature provides users with more creative possibilities.
Khan believes that in the future, companies like Spotify and YouTube should open up their data and provide APIs to promote the further development of the AI music industry.