We are on a mission to advance music technology

Innovation in technology knows no borders. It is a worldwide endeavor to address the intricate challenges of the music industry landscape. Our scientists and engineers embody this ethos with their diverse backgrounds and unique perspectives. We are a living proof that the collective creativity of a diverse team can compose symphonies of innovation.

The team

Join the researchers behind our mission

We're pioneering the future of music-tech with cutting-edge research. Be a part of our mission to empower creative potential through technology.

View current openings

Research Areas

Our primary objective is to foster innovation that has the potential to empower not just individual musicians but the entire music industry as a whole.

Music signal processing

This area focuses on the analysis and manipulation of music signals, including pitch detection, beat tracking, sections and melody extraction.

Source Separation

Source separation is a core feature that breaks down complex musical signals into their individual components, like separating vocals from instruments.

Music transcription

This involves the automatic conversion of music audio into a symbolic representation such as chords, sheet music or singing lyrics.

Voice synthesis

Transform the characteristics of one voice into another, maintaining the linguistic constructs during the conversion process.

Data and Engineering

Storing, organizing and processing huge datasets effectively.

New Trends

We're always staying on top of emerging technological advancements within the realm of music data science.

Recent Publications

Latest news and publications from our research team.

News

Music AI Partners with Audible Magic to Revolutionize Music Licensing for Film and TV

Music AI and Audible Magic have joined forces to deliver an innovative solution for film and TV production companies facing licensing challenges across territories. By combining Music AI’s industry-leading cinematic stem separation technology with Audible Magic’s gold-standard music identification, the partnership offers companies unprecedented control over content and licensing management.

News

Music AI Brings Industry-Leading Stem Separation to SourceAudio

Music AI is partnering with SourceAudio to bring unparalleled stem separation capabilities to the music and entertainment industry. Through this integration, SourceAudio’s 550,000 active users—ranging from broadcasters and media giants to individual artists—will gain access to Music AI’s cutting-edge stem separation technology, powered by state-of-the-art AI models. For anyone involved in audio production, this opens up a world of creative possibilities.

Research

Comparative Analysis Shows Music.AI Outperforms OpenAI in Lyric Transcription Accuracy

Music.AI outperforms OpenAI with a Word Error Rate that is 27.49% more accurate and a Character Error Rate that is 38% more accurate when paired with Music.AI’s complementary source separation technology.

Research

Music.AI Stem Separation Outperforms Commercial and Open-Source Alternatives

A new Signal to Distortion Ratio (SDR) benchmarking study confirms that Music.AI's model is uniquely designed for superior source separation, optimized to manage the dynamics of diverse audio elements with unmatched precision. While other solutions might struggle with complex audio mixes, Music.AI's approach sets a new standard in stem isolation for music production and analysis.

Eddie Hsu’s talk at the Web Summit Rio 2024 (Center Stage).

News

Web Summit Rio 2024: Music.AI’s COO Talks AI and Creativity

From an innovative idea to worldwide success, Music.AI’s journey is a testament to AI’s transformative impact on music and audio creativity. Read the highlights of the presentation by Eddie Hsu, our co-founder and COO, at the Web Summit Rio 2024.

News

Webinar: Matt Henninger Talks Revenue Growth and Music.AI

We collaborated with A2IM to discuss our platform and the use of AI to increase revenue in the music industry. Check out what Matt Henninger, our VP of Business Development, shared about our partnerships, voice synthesis, translation, and other topics.

Datasets

MoisesDB

MoisesDB is a comprehensive multitrack dataset for source separation beyond 4-stems, comprising 240 previously unreleased songs by 47 artists spanning twelve high-level genres. The total duration of the dataset is 14 hours, 24 minutes and 46 seconds, with an average recording length of 3:36 seconds. MoisesDB is offered free of charge for non-commercial research use only and includes baseline performance results for two publicly available source separation methods.

SDXDB23_LabelNoise

The purpose of this dataset is to provide the research community with a set of songs that can be used to design and evaluate source separation system under robust separation settings (i.e., when the ground truth data contains errors and inconsistencies). This dataset contains simulated errors regarding the identity of the musical instruments included in each track: we call these errors Label Noise.

SDXDB23_Bleeding

The purpose of this dataset is to provide the research community with a set of songs that can be used to design and evaluate source separation system under robust separation settings (i.e., when the ground truth data contains errors and inconsistencies). This dataset contains simulated errors for which each source is also present in the recording of all others: we call this phenomenon Bleeding.