While some would argue that the UK government should be trying to limit the role of AI in the music industry, its Creative Catalyst scheme actively seeks to advance it.
This comes via Innovate UK, part of UK Research and Innovation, and has resulted in the awarding of £1 million in funding to projects that “advance the development of Artificial Intelligence products and services within the global music supply chain which benefit and strengthen the UK Music Sector.”
It’s now been confirmed that one of the six ‘lead organisations’ to benefit from some of this money is AudioStrip, which is focused on “fine-grained music source separation with deep learning models”.
Assessors advised that "this is well planned, resourced and researched innovation that can impact the business, market and wider industry in the field of AI and music separation" and that "the rewards could be significant".
One of the stipulations of the funding is that those who receive it should “collaborate with industry stakeholders to unlock the full potential of AI applications across the value chain, supporting and uplifting the music industry, and driving long term creative and commercial success”. Going forward, AudioStrip will be working with C4DM Queen Mary University of London to create “state-of-the-art” AI music source separation.
Of course, there are several platforms out there that do this already, but AudioStrip says that it plans to go beyond those that can only separate vocals, bass and drums and create something that can identify and isolate more instruments at usable quality. In order to achieve this, it says it will develop advanced machine learning algorithms that can automatically detect musical instruments for high-quality audio source separation.
“This technology is sweeping the music industry,” says AudioStrip co-founder and CEO, Basil Woods. “AudioStrip will offer more advanced tools for precise separation of individual elements in audio files.
“By partnering with Queen Mary, we aim to elevate music source separation technology beyond industry benchmarks, making it an indispensable tool for DJs, independent artists, producers, and licensors.
“Our goal is to automatically identify musical elements from any given song - including vocal, instrumental, drums, bass, piano, electric guitar, acoustic guitar, and synthesizer - and extract them into independent tracks without losing quality.”
We're guessing that this is all for the future, but AudioStrip is already up and running and you can try it for free.
