arXiv:2506.14293v4 Announce Type: replace-cross Abstract: We present Sleeping-DISCO 9M, a large-scale pre-training dataset for music and song. To the best of our knowledge, there are no open-source high-quality dataset representing popular and well-known songs for generative music modeling tasks such as text-music, music-captioning, singing-voice synthesis, melody reconstruction and cross-model retrieval. Past contributions focused on isolated and constrained factors whose core perspective was to create synthetic or re-recorded music corpus (e.g. GTSinger, M4Singer) and arbitrarily large-scale
The increased maturity of generative AI models for various modalities is driving the need for corresponding high-quality, large-scale datasets to further advance capabilities.
Open-source, high-quality datasets for generative music modeling are crucial for accelerating innovation in AI-driven music creation and audio synthesis, democratizing access to powerful tools.
The availability of a 'popular and well-known' music dataset for training purposes suggests a significant leap towards more realistic and commercially viable AI-generated music and audio applications.
- · AI music startups
- · Generative AI researchers
- · Music producers leveraging AI tools
- · Creative industries
- · Proprietary music dataset providers
- · Traditional music composition software companies
- · Audio synthesis companies reliant on older models
Further rapid improvement in the quality and versatility of AI-generated music and audio.
New business models emerging around AI-assisted music creation, licensing, and personalization.
Potential for substantial disruption in the music industry's value chain, impacting artists, labels, and distribution.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG