AI Singapore launches competition to design solutions for detecting fake media
File photo of a man using a laptop. (Photo: AFP/Frederic J Brown)
SINGAPORE: A competition aimed at attracting the artificial intelligence (AI) community to design solutions that will help detect fake media was launched on Thursday (Jul 15).
The initiative, "Trusted Media Challenge", is a five-month-long competition driven by AI Singapore (AISG), a national AI programme launched by the National Research Foundation.
Participants will design and test out AI models and solutions that will easily detect audiovisual fake media - where both video and audio modalities may be modified.
The competition is targeted at AI enthusiasts and researchers around the world and aims to strengthen Singapore's position as a global AI hub by "incentivising involvement of international contributors, and sourcing innovation ideas globally", said AISG in a media release.
Participants will have access to datasets of original and fake media videos with audio.
The challenge is conducted in partnership with Mediacorp's CNA and Singapore Press Holdings' The Straits Times. The two have provided about 800 real video clips, including news and interviews. Custom videos were also collected from actors.
In total, about 4,000 real clips and 8,000 fake video clips will be made available for participants to train and test their models on.
READ: Deepfakes, nudes and the threat to national security
This challenge is open to researchers and industry professionals from all around the world, as well as anyone who is interested or experienced in machine learning, deep learning, and computer vision, especially in media forensics.
Participants will need to build AI models that "estimate" the probability that any given video is fake.
Mr Willy Tan, Lead of AI Strategy & Solutions at Mediacorp's News Group said it has "become easier" to create fake content that is "difficult for the human eye to differentiate", with deepfake technology becoming "more sophisticated and available".
"Maliciously doctored content can lead to public misinformation and social fissures, if left unchecked.
"CNA is excited to partner in the Trusted Media Challenge, collaborating in continued efforts to combat this impending threat, in our mission to provide timely and accurate news to Singapore and the region," said Mr Tan.
The Trusted Media Challenges opens on Jul 15, 2021. Interested participants can obtain full details and training data via the challenge platform provided by AISG at trustedmedia.aisingapore.org.
READ: Commentary: As information warfare enters a new age, time for a Total Defence upgrade
From Jul 15 to Dec 15, 2021, participating teams can submit their solutions - codes and models - via the challenge platform provided.
The platform will automatically rank the submissions on the leaderboard, said AISG.
The winner of this challenge stands to earn S$100,000 and a start-up grant of S$300,000 to develop their solutions further, using Singapore as the development base.
Prizes and start-up grants will be awarded to the top three winners, with the total prize money amounting to S$700,000.
The challenge is divided into two phases, with Phase 1 lasting four months and the top teams from Phase 1 will enter Phase 2.
Prize money will be given out based on the ranking in Phase 2, said AISG. The announcement of the top three winners is expected to take place in January 2022.
