China’s research and technology sector is increasingly exploring AI systems that can recognize and decode human emotions. A recent project led by Zhejiang University researcher Xu Xin used a dynamic emotion-recognition system to screen students at a middle school in Weihai City, Shandong Province, reportedly measuring 14 emotional indicators, including stress, anxiety, and depression. The system can reportedly detect subtle physical movements with 0.01-millimeter precision and achieve 90.8 percent emotion-recognition accuracy.
Zhejiang University has also developed AI-based approaches to emotional and psychiatric assessment. In 2025, teams from the university jointly developed the Emoface model, which analyzes detailed facial features to distinguish major depressive disorder from bipolar disorder. Another Zhejiang University team published research in July 2026 on emotion decoding using an implanted brain-computer interface. The article describes these efforts as part of a broader push to quantify the links between brain activity, emotions, and behavior.
Growing demand for mental-health services, advances in multimodal AI, and the development of wearable sensors are creating opportunities for emotion AI. Major technology companies, including Huawei, Google, Microsoft, and Amazon, have explored emotion- or sentiment-related technologies, while startups are developing AI companions and emotional-support toys. The global emotional-AI market is projected to grow substantially over the coming decade.
Source: 36kr.com, September 9, 2026
https://www.36kr.com/p/3976937208158087