Monash University
The University of Queensland
MBZUAI
American University of Sharjah
After two years successfully hosting 1M Deepfake detection challenge at ACM Multimedia 2024, 2025 with more than 200+ teams participating in the challenge,
this year our focus is more on the explainability part of the manipulated content, along with the detection performance.
Task 1: Deepfake Detection – Given an image sample, the task is to identify if the image is a deepfake or real.
Task 2: Deepfakes Explainability – Given an image sample, the task is to find out the explanation behind the model’s decision with respect to the manipulation. The assumption here is that from the perspective of spreading misinformation.
Step 1: For participating in the challenge, the first step is to register your team. Please complete the form (coming soon) with the required information.
Step 2: The second stage is the EULA agreement (coming soon). Please download the dataset EULA and sign it at your end. Please email the signed EULA to deepfakes1mchallenge@gmail.com
For details, please contact Abhinav Dhall.
Last updated: 12 Mar 2026
Website credit: This website style is adopted from the tutorial page.
