Welcome to the AVI Challenge 2026!
Assessing True Personality Traits and Cognitive Ability from Asynchronous Video Interviews
Call for Participation: AVI Challenge 2026
The 34th ACM International Conference on Multimedia Grand Challenge invites you to join the AVI Challenge 2026: Assessing True Personality Traits and Cognitive Ability from Asynchronous Video Interviews!
Why Join?
Asynchronous Video Interviews (AVIs) are revolutionizing hiring, leveraging AI to assess personality traits and cognitive ability. However, past challenges have faced limitations in methodology and validity. AVI Challenge 2026 is here to change that!
What Makes This Challenge Unique?
• Realistic & Structured: Our dataset features simulated job interviews with carefully crafted questions based on Trait Activation Theory (TAT).
• High-Quality Annotations: Trained annotators and professional recruiters ensure accurate personality and competency ratings using Behavioral Response Rating Scales (BARS).
• Practical Impact: The dataset mirrors real AVI-based hiring, making it directly applicable to industry needs.
Be Part of the Future!
This is your chance to push the boundaries of AI-driven hiring assessments, benchmark your models, and contribute to fair, reliable talent evaluation. Plus, we’re expanding our dataset to include multilingual responses, driving global research forward!
Join us at AVI Challenge 2026 and make an impact—we can’t wait to see what you create!
Results Announcement
We are pleased to announce the final results of the AVI Challenge 2026. Congratulations to the top three teams, who are granted paper recommendation eligibility for the ACMMM 2026 Main Track.
| Overall Rank | Team Name | Track 1 (MSE) | Track 1 Rank | Track 2 (ACC) | Track 2 Rank |
|---|---|---|---|---|---|
| 1 | avi_team | 0.2777 | 1 | 0.7923 | 4 |
| 2 | personax | 0.3633 | 19 | 0.8692 | 1 |
| 3 | visionxl | 0.2780 | 3 | 0.8077 | 3 |
The three teams above are kindly requested to prepare their manuscripts and submit them by June 7 to qitianhua@seu.edu.cn.
The author list must remain identical to the registration form. Only the order of authors may be adjusted. Adding or removing authors is strictly prohibited.
Please ensure the submission meets ACM Multimedia paper quality standards. Submissions that clearly violate ACM MM requirements or involve manipulation of experimental results will be rejected.
We sincerely thank all participating teams for their strong support. The AVI dataset is expected to be released in H2 2026 / H1 2027. Teams considering alternative submissions are encouraged to stay tuned.