2025 Automatic Music Transcription Challenge

Schedule (tentative)

Cash Awards

• The top award is up to $1,500 US Dollars.
• The second award is up to $1,000 US Dollars.
• The third award is $500 US Dollars.

The winners must open source their solutions before receiving the cash awards (required in the registration agreement). The source code should have proper documentation for reproducibility. The organizers will examine the source code before announcing the winners. The cash awards can be given to only the participants that are not in the embargoed and sanctioned countries declared by the United States.

The award may be in the form of cash or travel grants that may include conference registration + hotel + airplane ticket.

Sample Compositions

We are releasing 20 sample compositions to provide participants with a clear understanding of the type of music their models will be graded on. These compositions showcase a variety of instrumental arrangements, helping participants fine-tune their models to produce high-quality results.

The included instruments are:

1. Piano
2. Violin
3. Cello
4. Flute
5. Bassoon
6. Trombone
7. Oboe
8. Viola

Each composition is accompanied by an MP3 audio file and its corresponding sheet music, available in both MIDI and PDF formats.

You can access the sample compositions here: Sample Compositions Google Drive Link

Sample Solution

As a reference implementation, we present MT3 (Multi-Task Multitrack Music Transcription) developed by Google's Magenta team. MT3 is a state-of-the-art model designed to transcribe music involving multiple instruments simultaneously. It utilizes a Transformer-based architecture to process audio inputs and generate corresponding musical notations.

For a comprehensive understanding of MT3's architecture and implementation, you can refer to the following resources:

• Research Paper: MT3: Multi-Task Multitrack Music Transcription
• GitHub Repository: magenta/mt3

Participants are encouraged to study MT3 as a foundational model to gain insights into effective music transcription architectures. It serves as a valuable starting point for developing and refining your own transcription models.

Note: To be considered for a winning submission, your model must achieve a higher transcription accuracy score compared to MT3. This benchmark ensures that new models contribute meaningful advancements to the field of music transcription.

Organizers

Photo	Name	Email	Organization
	Kristen Yeon-Ji Yun	yun98@purdue.edu	Department of Music, Purdue University
	Yung-Hsiang Lu	yunglu@purdue.edu	School of Electrical and Computer Engineering, Purdue University
	George K. Thiruvathukal	gthiruvathukal@luc.edu	Department of Computer Science, Loyola University Chicago
	Tae Hong Park	thp@purdue.edu	Department of Music, Purdue University
	Harry Bulow	hbulow@purdue.edu	Department of Music, Purdue University
	Allen McCullough	mccullo3@purdue.edu	Department of Music, Purdue University
	Ojas Chaturvedi	ochaturv@purdue.edu	Department of Computer Science, Purdue University

Contributing Composers

Photo	Name	Email	Organization
	Harry Bulow	hbulow@purdue.edu	Department of Music, Purdue University
	Tae Hong Park	thp@purdue.edu	Department of Music, Purdue University
	Allen McCullough	mccullo3@purdue.edu	Department of Music, Purdue University
	Hubert Howe	hubert.howe@gmail.com	Juilliard School of Music (retired)
	Ka-Wai Yu	ka-wai.yu@utahtech.edu	Department of Music, Utah Tech University
	Monte Taylor	mtaylor@purdue.edu	Department of Music, Purdue University
	Mika Pelo	mpelo@ucdavis.edu	University of California, Davis

For Contributing Composers

For the 2025 challenge, the composers are by invitation by the organizers only. The reason is that the organizers wish to manage the difficulty of the music so that computer programs can provide meaningful results. If a composer is interested in contributing, please contact the organizers. The composers will retain their copyright. Composers will agree to provide their music royalty-free and non-exclusive rights to the challenge organizers for the purpose of this automatic music transcription challenge, including (but not limited to) redistributing the music as samples and analyzing the music by computer programs. The organizers may modify the music from the composers to meet the restrictions of the scoring program. The names, profile photos, and links to the composers' websites will be listed on the https://ai4musicians.org/ website if a composer is selected by the organizers.

A composer will provide at least 10 and at most 30 pieces of music. Each piece is approximately 20 seconds long. These numbers aim to give each composer a reasonable representation without causing computer programs to "overfit" into a specific composer's style. Each piece is between 10 and 20 seconds. The music should be divided into 3 levels of difficulties: easy, medium, and difficult.

• Tempo range from 60 bpm-90 bpm
• Pitch range should be from C2 to C7
• The smallest rhythmic duration should be sixteenth-notes or sixteenth-rests
• No interpreted rhythms such as swing feel patterns (must ne clearly discernable without interpretation) If desiring a swing pattern, write as dotted eighth and sixteenth-note or quarter-note/eight-note triplets
• No doubly-dotted notes
• Can be rhythmically and melodically diverse
• Stay within the following meters: 3/4, 4/4, 6/8.
• No trills or mordents of any type.
• Keep the dynamic range from pianissimo to fortissimo (pp, p, mp, mf, f, ff)
• Tonal, modal and atonal styles, or a blend of these, are all encouraged
• Up to three different instruments may be used in each composition. Avoid using violin and cello in combination with each other, due to their common timbral characteristics. When writing for multiple instruments (limit of 3), use a diversity of timbre combinations such as (1) flute and trombone, (2) violin and bassoon, (3) cello, oboe, and bassoon. Any combination of instruments is fine as long as the timbre of each is clearly discernable.
• Submit a score (PDF), MusicXML, and MIDI recording of each work.

List of Instruments

1. Piano
2. Violin
3. Cello
4. Flute
5. Bassoon
6. Trombone
7. Oboe
8. Viola

The instrument groups should not include instruments that can be misclassified easily even by human experts (e.g., violin and viola).

Frequently Asked Questions