TRUSTED MEDIA CHALLENGE
BACKGROUND
Fake media is an existential threat to societies today. If left unchecked, it risks becoming a serious national security concern. AI Singapore is launching the Trusted Media Challenge to test out solutions and explore how Artificial Intelligence (AI) technologies can be leveraged to combat fake media.
TASK
Given a set of original and fake videos, the participants will be asked to build models to return a probability that any given video is a fake one.
DATASET
The challenge dataset consists of both real and fake media (i.e. videos with audio). The real videos are from various sources including news reporters and consented actors, with the majority of Asian ethnicity. The fake media are created using different fake media generation methods based on the real media described above. The videos in the dataset have various durations, background, illumination, a minimum resolution of 360p and may contain perturbations that mimic transmission errors and bad compression. Perturbations exist in both real and fake media and such perturbations are not regarded as manipulation.
[The TMC training dataset is available post-challenge to the research community for further research. Please contact us if you are interested.]
TIMELINE
15 Nov 2021
PHASE 1:END DATE
15 Dec 2021
PHASE 2:END DATE
15 Jul 2021
PHASE 1:START DATE
16 Nov 2021
PHASE 2:START DATE
Jan 2022
ANNOUNCEMENT OF WINNERS
PRIZE
The top three teams will be declared as the winners of the Challenge. They will receive both a cash prize and a start-up grant, less applicable taxes, if any. Please refer to the following table for the detailed information:
Rank Cash prize (SGD) Start-up grant (SGD)
1 100,000 300,000
2 50,000 150,000
3 25,000 75,000