The field of affective computing is moving towards more nuanced and comprehensive understanding of human emotions, with a focus on real-world applications and longitudinal studies. Recent research has highlighted the importance of considering emotional dynamics and stress in various settings, including workplaces and call centers. Innovative methods for micro-expression recognition and emotional state estimation have been proposed, leveraging advances in deep learning and computer vision. These developments have the potential to improve emotion-aware system design, employee performance prediction, and customer service quality. Noteworthy papers include:

• WELD, a large-scale longitudinal dataset of emotional dynamics, which enables research in emotion recognition and affective dynamics modeling.
• Improving Micro-Expression Recognition with Phase-Aware Temporal Augmentation, which proposes a novel augmentation method for recognizing subtle facial transitions.
• Call-Center Staff Scheduling Considering Performance Evolution under Emotional Stress, which presents a memetic optimization algorithm for scheduling call-center staff based on emotional stress and performance evolution.