SHANGHAI, Sept. 6, 2024 /PRNewswire/ — The Global Multimedia Deepfake Detection Challenge 2024(The Challenge) has successfully concluded with over 2,200 participants from 26 countries. Co-hosted by the INCLUSION Conference and Ant Group, and organized by Ant Digital Technologies, the event showcased the potential of technology when leveraged for societal benefit.
Initiated in July 2024, the Challenge aims to invite participants to develop, test, and further evolve more accurate, effective, and innovative detection models against various types of Deepfake attacks in real-world scenarios, as well as to motivate innovative defense strategies and improve Deepfake recognition accuracy. The Challenge is structured into two tracks: an image track and an audio/video track, featuring datasets with multi-dimensional face forgery methods.
In the Image Track, a collaborative team from Hong Kong and Macau, led by Dr. Wu Haiwei of City University of Hong Kong, achieved an accuracy rate of 97.038% in detecting deepfakes across various image formats. In the Audio/Video Track, individual competitor Tang Yongwei dominated with cutting-edge algorithms for effectively identifying manipulated audio.
“The Challenge provides researchers with a highly simulated industrial environment for practice, helping to integrate industry, academia, and research while cultivating responsible practical talents,” said JOEY Zhou Tianyi, Deputy Director of the Centre for Frontier AI Research (CFAR) at the Agency for Science, Technology and Research in Singapore.
William Yao, CTO of ZOLOZ and a core organizer of the Challenge, emphasized the importance of fostering awareness around the positive uses of technology. “We hope to improve the industry’s security defense level with the involvement of multiple parties; this is also the original intention behind hosting the challenge,” Yao remarked.
On August 30th, the organizing committee launched an open-source initiative: “After participants voluntarily open source their work, the organizing committee will collect all open-source repositories on their official GitHub homepage.” The initiative was quickly met with positive responses from the VisionRush team of the Institute of Automation of the Chinese Academy of Sciences and the JTGroup from the University of Macau. So far, nearly 100 participants and teams of the Challenge have expressed their willingness to actively participate, hoping to protect people from deepfakes.
About The INCLUSION•Conference on the Bund
The inaugural INCLUSION Conference on the Bund successfully concluded on September 26, 2020, with the guidance of the Shanghai Municipal People’s Government and the support of the Shanghai Municipal Financial Regulatory Bureau and Shanghai Huangpu District People’s Government. The 2023 INCLUSION Conference on the Bund included an 8,000 square meter technology exhibition and a 3,000 square meter green marketplace for conference goers from around the world.
For more details of the upcoming conference, visit the official website: inclusionconf.com
View original content to download multimedia:https://www.prnewswire.com/apac/news-releases/2024-global-multimedia-deepfake-detection-challenge-winner-announced-and-exploring-the-future-of-responsible-technology-302240478.html
SOURCE Ant Digital Technologies
