Privacy focused challenge tests AI for teen depression and anxiety detection
AdoDAS: A Privacy-Preserving Multimodal Challenge for Adolescent Depression, Anxiety, and Stress Assessment
MultimediaSound
Summary
Mental health issues like depression, anxiety, and stress among teenagers need tools that help doctors without replacing them. The authors created a contest where teams used safe, anonymous audio and video data to detect these conditions. This contest included over 6,000 young participants answering questions while keeping their private data secure. The best computer systems used clever methods to combine different types of information over time to better predict mental health symptoms.
What this means in practice
- •For mental health software developers: Create tools that assess teen depression and anxiety from anonymized audio-visual data without exposing private raw recordings.$Commercial implications: Enables privacy-respecting mental health assessment apps for schools or clinics, addressing legal and ethical data concerns.
- •For digital health platform teams: Integrate multimodal, session-aware models to improve automated screening accuracy for adolescent mental health conditions.
Authors
Zhaojie Luo, Junkun Wang, Tianhua Qi, Yuxuan Wu, Xin Zhao, Tetsuya Takiguchi, Tomoko Matsui, Kun Qian, Fei Wang, Shuqiong Wu, Zhengjun Yue, Hiroshi Ishiguro, Xinyuan Qian, Haizhou Li
Abstract
Adolescent depression, anxiety, and stress (D/A/S) call for scalable tools that complement, rather than replace, professional evaluation. Under a privacy-preserving policy, the AdoDAS Grand Challenge withholds minors' raw recordings and distributes anonymized audio-visual representations and ASR-derived text. Its 6,000 participants provide 24,000 segments across one scripted-reading and three open-response sessions. Two tracks assess multi-task binary D/A/S screening and ordinal prediction of 21 DASS-21 item responses. From 191 registrations, the final leaderboards included 95 eligible screening teams and 64 item-prediction teams. Audio-visual baselines achieved 0.4604 mean F1 and 0.2675 mean Quadratic Weighted Kappa; leading submissions reached 0.5921 and 0.2776. Representative systems emphasize cross-session modelling, temporal multimodal fusion, psychometric structure, and task-aware calibration.