Call for Papers | 前沿情感计算技术及其神经机制：从情感计算到定量神经认知建模

Virtual Collection（VC，虚拟专辑）是推动学科开放学术交流的创新出版形式。区别于传统专刊，VC 打破传统专刊模式的固有局限，围绕特定主题开展定向征稿与专题归集，鼓励全球科研人员以更灵活、多元的方式开展学术合作，共同搭建高水平学术交流与资源共享平台。

This Virtual Collection focuses on the synergistic integration of advanced affective computing, quantitative computational approaches, and neural mechanisms of emotion. As affective computing technologies advance from pattern recognition toward biologically grounded, interpretable, and clinically actionable systems, a central challenge is to bridge the gap between machine learning paradigms and the mechanistic understanding of emotional processes in the human brain. The collection will connect computer science, computational neuroscience, cognitive psychology, information theory, and biomedical engineering, prioritizing research that translates neural findings into computable affective models, validated against physiological and behavioral evidence.

It will particularly welcome studies that move beyond performance-oriented benchmarks to quantify neural interpretability, evaluate bio-constraint validity, and identify enabling theoretical frameworks for next-generation emotional AI. Topics include multimodal affective modeling integrating Stereoelectroencephalography (SEEG), functional magnetic resonance imaging (fMRI), magnetoencephalography (MEG), wearable physiological sensors, and behavioral cues; deep learning architectures with neurobiologically plausible components and explainability mechanisms grounded in established brain circuitry; computational accounts of emotion dynamics, self-regulation, and adaptation informed by reinforcement learning, Bayesian inference, and dynamic systems theory; information-theoretic and network-science approaches to quantifying emotional complexity, uncertainty, and interpersonal synchronization; translational applications in precision mental health, including early detection, predictive intervention, and closed-loop neuromodulation; cross-cultural, developmental, and individual-difference factors modulating the neural basis of emotion; and affective brain-computer interfaces enabling real-time emotion decoding, neurofeedback, and adaptive human-machine interaction.

The Virtual Collection welcomes original research articles, systematic reviews, methodological advances, perspective pieces, and data papers that provide open, reproducible, and clinically or technologically actionable evidence. Contributions integrating empirical neuroscience with quantitative computational modeling are particularly encouraged. Together, the articles should build an integrated foundation for neuro-informed affective computing, establish best practices for cross-modal validation, and chart a forward-looking research agenda at the intersection of artificial intelligence, cognitive neuroscience, and emotional computation.

It will particularly welcome studies that move beyond performance-oriented benchmarks to quantify neural interpretability, evaluate bio-constraint validity, and identify enabling theoretical frameworks for next-generation emotional AI. Topics include but not limit:

[T1] Multimodal Affective Modeling and Neurophysiological Signal Integration

Integration of Stereoelectroencephalography (SEEG), functional magnetic resonance imaging (fMRI), magnetoencephalography (MEG), wearable physiological sensors (electrodermal activity, cardiac rhythm, respiration), eye-tracking, and behavioral cues to explore spatiotemporal dynamics of emotional states; development of unified representation learning frameworks using graph neural networks, variational autoencoders, and transformer-based architectures for multi-scale signal fusion and cross-modal validation.

[T2] Deep Learning and Explainable Neural Decoding

Advanced convolutional, recurrent, transformer and LLMs-based emotion recognition models with neurobiologically plausible components; interpretability mechanisms including saliency maps, adversarial perturbations, and causal inference approaches to reveal correspondences between model representations and neural circuits such as the amygdala, prefrontal cortex, and limbic networks; explanatory validation against established findings in affective neuroscience.

[T3] Computational Neural Mechanisms of Emotion Regulation

Computational models of emotional self-regulation grounded in reinforcement learning, Bayesian optimization, and dynamic systems theory; integration with neuromodulation experimental data (tDCS, TMS, TI et al.) to validate predictive accuracy of emotion regulation effects; exploration of computational psychiatry models linking regulatory dysfunction to affective disorders.

[T4] Statistical and Information-Theoretic Frameworks for Quantitative Affective Computing

Information-theoretic and network-science approaches to quantifying emotional complexity, uncertainty, and interpersonal synchronization using entropy, mutual information, Bayesian inference, and network entropy metrics; critical phase transition models for emotional state transitions; development of robust statistical frameworks for measuring affective dynamics across individuals and contexts.

[T5] Precision Prediction and Intervention in Mental Health through Affective Computing

Construction of affective biomarkers using wearable devices, mobile sensing, and clinical assessments for precision mental health applications; predictive models for early detection of depression, anxiety, and related conditions with neural mechanism validation; development of personalized intervention frameworks that bridge computational modeling and clinical implementation.

[T6] Neural Basis of Emotion across Cultures and Individual Differences

Investigation of how cultural, educational, gender-related, genetic, and developmental factors modulate emotional brain networks; development of cross-situational, cross-population affective computing models that enhance robustness and fairness; comparative neuroimaging studies examining universality and cultural specificity in emotional processing.

[T7] Affective Brain-Computer Interfaces and Closed-Loop Neurofeedback

Real-time emotion decoding systems integrated with brain stimulation or robotic devices for closed-loop emotional regulation; evaluation of safety, efficacy, and neuroplasticity changes in closed-loop architectures; development of trustworthy affective brain-computer interfaces (ABCIs) with adaptive human-machine interaction capabilities and robust feedback mechanisms.

insights@the-innovation.org