Deadline: 18 June 2025
The AI Singapore and the Institute for Information & Communication Technology Planning & Evaluation of South Korea are jointly launching a grant call to provide funding to support research in AI for Deepfake Detection.
The aim of this joint grant call is to support the development of scientific and technological cooperation between Singapore and South Korean researchers and to strengthen the scientific partnership between research groups from both countries, by establishing bilateral research networks, enhancing research cooperation, and promoting the exchange of knowledge between the researchers of the two countries, in the field of artificial intelligence.
This joint grant call focuses on developing improved detection methods capable of counteracting evolving deepfake threats, enabling users to detect manipulated and/or AI generated content that understands different linguistic, socio-cultural, and environmental characteristics.
Topic
- The joint grant call is to support Research for High Real-World Accuracy Deepfake Detection.
Areas of Research
- The research should encompass (but is not limited to):
- Research and development of novel, high-accuracy, scalable, multimodal AI deepfake detection technology that identifies deepfake manipulation in video, voice and text inputs, and provides intuitive explanations of authenticity using bounding boxes, captions, confidence scores etc. to help users easily understand the manipulated parts.
- Application of the deepfake detection technologies to at least 3 distinctly different industry use cases (e.g., social media, news outlets, voice biometrics etc.) relevant to Singapore and/or South Korea context, with each application supported by at least one or more industry and/or agency partners from each country committed to using and/or commercializing the technology.
- Jointly building large-scale datasets with national languages, dialects, and sociocultural and regional environmental characteristics, as well as developing the required data pipelines to train and maintain deepfake detection model(s).
- Benchmarking and validation of deepfake detection technology with prevailing state-of-the-art generative AI models, as well as with industry datasets relevant to the application use case proposed by the research team.
- Deriving requirements, designing and implementing architecture, functionality, and interfaces for a global common platform for deepfake detection
- Prototyping a cloud-based open global common platform for deepfake detection that can be jointly used by multiple countries.
Funding Information
- The joint grant call will award only one joint research team consisting of researchers from both Singapore and South Korea.
- The Singapore research team will receive a grant quantum of up to SGD1,500,000 from AISG, to support activities conducted in Singapore, for up to three years.
- The South Korea research team will receive a grant quantum of up to KRW1,600,000,000 from IITP, to support activities conducted in South Korea, for the same duration.
Expected Outcomes
- The desired outcomes are:
- A functioning minimum viable product (MVP) with design performance specification, which is jointly developed, validated and test-bedded by Singapore and South Korea researchers.
- Joint publication of paper(s) with researchers from both Singapore and South Korea.
Eligibility Criteria
- In Singapore, the grant call is open to the six Singapore-based Institutes of Higher Learning (IHLs) and A*STAR Research Institutes. Proposals are invited from multidisciplinary teams and consortia of academics, researchers, scientists, engineers, domain experts, and other professionals.
- In South Korea, the grant call is open to any South Korea-based organization, but the consortium must include organizations capable of operating platforms.
- Each joint team must be led by one Principal Investigator (PI) from Singapore and one Principal Investigator from South Korea. The Singapore team must comprise one PI and at least one Co-PI.
For more information, visit AISingapore.