Meta researchers introduce Audiobox, a new foundation audio generative model with enhanced controllability that aids users to more easily create their desired voices and sound effects. Audiobox can generate speech, voices and sound effects using a combination of voice inputs and natural language text – making it easy to create custom audio for a wide range of use cases.
The FAIR (Fundamental AI Research) team at Meta is interested to provide:
Access to Audiobox models and relevant code artifacts to groups of researchers at academic institutions and research labs
Funding up to $50,000 to further their work.
Request for Proposals
At Meta, we deeply care about conducting audio and speech research responsibly and have implemented new technologies towards that end. We also believe issues in audio and speech research cannot be addressed by an individual or single organization alone. In particular, we value a diverse set of partners to tackle the Responsible AI aspects of this work. We are launching a call for proposals from two groups:
Researchers who want to conduct Responsible AI Research and advance areas such as safety, fairness, ethics and impersonation / deepfake detection, and enable a future where speech and audio models can be used responsibly and safely.
Researchers who want to advance state-of-the-art speech and audio generation performance by leveraging AudioBox models.
To foster further innovation and to deepen our work in audio generation, we invite academic and research organizations with initiatives focused on Responsible AI, Safety, Ethics, and Audio/Speech research. We anticipate selecting up to 10 partner institutions and providing up to $50,000 each in grants to further their work.
Grant payments will be made to the host institution per the Terms and Conditions set forth below. We strongly encourage organizations and researchers who want to focus on Responsible AI and areas in safety, fairness, ethics and impersonation / deepfake detection to apply. A strong application should include a clear research agenda, the impact of the project, a plan with laid out milestones, and how the budget will be used. Recipients will receive access to Audiobox models with relevant code artifacts and funding as an unrestricted gift.
Audiobox represents a huge step toward democratizing audio generation capabilities. Audio plays a fundamental role in many forms of media, from movies to podcasts, audiobooks, and video games. In the future, we believe research breakthroughs such as these will lower the barrier of accessibility for audio creation and make it easy for anyone to become an audio content creator.
Application Dates & Deadlines
Launch date: December 11, 2023
Deadline: February 2, 2024 at 11:59pm PST
Estimated notification date: March 8, 2024
Budget and Payment
Award amount can vary but may be up to $50,000 each for up 10 grant recipients
Payment: The award will be made to the PI’s host institution.
Proposal Requirements
A summary of the project, in English:
Research agenda
Anticipated impact of project
Any relevant prior work
A timeline with milestones and expected outcomes (with 1 year).
We will invite recipients at the end of the 1 year term (March 2025) to share their research outcomes. Travel expenses should be accounted for as a part of your cash grant.
A budget description including an explanation of how funds would be spent
Curriculum Vitae for all project participants.
Organization details; this will include tax information and administrative contact details
Judging and Selection
A strong application should include a clear research agenda, the impact of the project (including how it will benefit the community), a plan with laid out milestones, and how the budget will be used. The plan should also state how the team will utilize the Audiobox models that you’ll be provided access to. While not required, we strongly value organizations and researchers who want to focus on Responsible AI.
Selection will be processed by the AI at Meta Research team.
Eligibility
Organizations must be an accredited academic institution, a research institution, or a non-governmental organization with recognized legal status in their respective country (equal to 501(c)(3) status under the United States Internal Revenue Code).
Meta cannot consider proposals submitted, prepared or to be carried out by individuals residing in, or affiliated with an academic institution located in, a country or territory subject to comprehensive U.S. trade sanctions.
Applicants must be the Principal Investigator (PI) on any resulting award. All funding will be directed to the PI’s institution.
Applicants may submit one proposal per solicitation (current) Applicants may submit individually or in teams, however, each applicant may only submit one proposal per solicitation.
Awards must comply with applicable U.S. and international laws, regulations and policies.
Application is not open to institutions in Texas and Illinois.
Terms and Conditions
Please review the program’s Terms and Conditions.
Audiobox Responsible Generation Grant
Meta researchers introduce Audiobox, a new foundation audio generative model with enhanced controllability that aids users to more easily create their desired voices and sound effects. Audiobox can generate speech, voices and sound effects using a combination of voice inputs and natural language text – making it easy to create custom audio for a wide range of use cases.
The FAIR (Fundamental AI Research) team at Meta is interested to provide:
Access to Audiobox models and relevant code artifacts to groups of researchers at academic institutions and research labs
Funding up to $50,000 to further their work.
Request for Proposals
At Meta, we deeply care about conducting audio and speech research responsibly and have implemented new technologies towards that end. We also believe issues in audio and speech research cannot be addressed by an individual or single organization alone. In particular, we value a diverse set of partners to tackle the Responsible AI aspects of this work. We are launching a call for proposals from two groups:
Researchers who want to conduct Responsible AI Research and advance areas such as safety, fairness, ethics and impersonation / deepfake detection, and enable a future where speech and audio models can be used responsibly and safely.
Researchers who want to advance state-of-the-art speech and audio generation performance by leveraging AudioBox models.
To foster further innovation and to deepen our work in audio generation, we invite academic and research organizations with initiatives focused on Responsible AI, Safety, Ethics, and Audio/Speech research. We anticipate selecting up to 10 partner institutions and providing up to $50,000 each in grants to further their work.
Grant payments will be made to the host institution per the Terms and Conditions set forth below. We strongly encourage organizations and researchers who want to focus on Responsible AI and areas in safety, fairness, ethics and impersonation / deepfake detection to apply. A strong application should include a clear research agenda, the impact of the project, a plan with laid out milestones, and how the budget will be used. Recipients will receive access to Audiobox models with relevant code artifacts and funding as an unrestricted gift.
Audiobox represents a huge step toward democratizing audio generation capabilities. Audio plays a fundamental role in many forms of media, from movies to podcasts, audiobooks, and video games. In the future, we believe research breakthroughs such as these will lower the barrier of accessibility for audio creation and make it easy for anyone to become an audio content creator.
Application Dates & Deadlines
Launch date: December 11, 2023
Deadline: February 2, 2024 at 11:59pm PST
Estimated notification date: March 8, 2024
Budget and Payment
Award amount can vary but may be up to $50,000 each for up 10 grant recipients
Payment: The award will be made to the PI’s host institution.
Proposal Requirements
A summary of the project, in English:
Research agenda
Anticipated impact of project
Any relevant prior work
A timeline with milestones and expected outcomes (with 1 year).
We will invite recipients at the end of the 1 year term (March 2025) to share their research outcomes. Travel expenses should be accounted for as a part of your cash grant.
A budget description including an explanation of how funds would be spent
Curriculum Vitae for all project participants.
Organization details; this will include tax information and administrative contact details
Judging and Selection
A strong application should include a clear research agenda, the impact of the project (including how it will benefit the community), a plan with laid out milestones, and how the budget will be used. The plan should also state how the team will utilize the Audiobox models that you’ll be provided access to. While not required, we strongly value organizations and researchers who want to focus on Responsible AI.
Selection will be processed by the AI at Meta Research team.
Eligibility
Organizations must be an accredited academic institution, a research institution, or a non-governmental organization with recognized legal status in their respective country (equal to 501(c)(3) status under the United States Internal Revenue Code).
Meta cannot consider proposals submitted, prepared or to be carried out by individuals residing in, or affiliated with an academic institution located in, a country or territory subject to comprehensive U.S. trade sanctions.
Applicants must be the Principal Investigator (PI) on any resulting award. All funding will be directed to the PI’s institution.
Applicants may submit one proposal per solicitation (current) Applicants may submit individually or in teams, however, each applicant may only submit one proposal per solicitation.
Awards must comply with applicable U.S. and international laws, regulations and policies.
Application is not open to institutions in Texas and Illinois.
Terms and Conditions
Please review the program’s Terms and Conditions.