Overview

With the increasing adoption of facial recognition systems in security, banking, and authentication applications, the need for reliable face anti-spoofing solutions has become critical. This challenge challenges participants to create models that can effectively detect and prevent presentation attacks while maintaining high accuracy for legitimate users

Challenge Objective

The Face Anti-spoofing Challenge aims to develop robust and accurate models that can distinguish between real human faces and various types of spoofing attempts, including:

• Printed photos
• Digital displays
• 3D masks
• Deepfake videos
• Other emerging spoofing techniques

Codebase and Data

• Model Codebase URL
• Challenge Dataset URL

Evaluation

Type of Metric

Models will be evaluated based on Accuracy metric.

Submission file

Format:

The submission file has two field:

• id: The unique name corresponding to each image
• label: The predicted label for the corresponding image

Example:

id,label
ce6e440e-2cfc-4684-97d7-fa3a7db7d799.png,Spoof
dec0a17f-9981-423e-8e0e-ebee7cb51b23.png,Real

Challenge Timeline

Challenge Phase	Duration	Start Date	End Date
Registration	2 weeks	25/06/2025	08/07/2025
Training Phase	1 week	09/07/2025	15/07/2025
Validation Phase	2 weeks	16/07/2025	29/07/2025
Final Submission	1 week	30/07/2025	05/08/2025
Judging Phase	2 weeks	06/08/2025	19/08/2025
Result Announcement	1 day	20/08/2025	20/08/2025

• Note: Timeline EST

Change log

🗓️ Public Dataset (for public eval)

• ✅ Dataset Phase 1: 09/07/2025
• ✅ Dataset Phase 2: 16/07/2025
• ✅ Dataset Phase 3: 29/07/2025

🗓️ Final Submission

• ✅ Upload your private solution: 30/07/2025

Prizes

• Total Challenge Rewards: 4,500 AIOZ tokens
• Special Reward: 2,000 AIOZ tokens
• Top 5 Participants: 500 AIOZ tokens each