BADEN-BADEN, Germany - June 9, 2020 - (Newswire.com)
Open Media LLC announces the launch of Lalal.ai, an innovative service extracting vocals and instrumentals from music tracks. Lalal.ai is powered by a machine-learning algorithm which accurately recognizes and separates voice and backtracks from audio files.
For training of the Lalal.ai neural network, the developers have used the music database of studio-quality multi-track recordings and compared the results with other audio splitting methods.
To help DJs, sound producers, vocalists, dancers, and karaoke singers to work with better-quality sounds, Open Media LLC engineers have used 20 TB of training data. 45 million neural network parameters and advanced digital signal processing have been used to separate vocals and instrumental tracks. Thus an advanced audio splitting method has been produced.
With Lalal.ai, users can extract voices and backtracks from any song, movie, podcast or video and download extracted audio in the same format and quality as the source file. The service supports all audio file formats.
The functionality of the Lalal.ai audio splitting tool is available online now and is free to use.
Lalal.ai Media Kit
Press Release Service by Newswire.com
Original Source: Lalal.ai Audio Sevice Revolutionizes AI-Powered Vocal and Instrumental Extraction