6 results about "Emotional Changes" patented technology

This invention relates to the field of EEG emotion recognition technology, specifically to an emotion recognition method and system based on two-level KL divergence. The method includes: acquiring EEG data and calculating the first KL divergence between the left and right hemispheres in each region before and after external sensory stimulation, at each set frequency band; averaging the first KL divergence of each segment in each hemisphere as the feature before and after external sensory stimulation; calculating the second KL divergence based on this feature; and concatenating the feature before and after external sensory stimulation with the second KL divergence as input to a classification model. This invention deeply mines the emotionally sensitive features in EEG signals through two-level progressive distribution difference calculation and combines it with a classification network to achieve model training and accurate classification, thereby achieving precise recognition of subtle emotional changes.

This invention discloses an emotional intervention system integrating multi-sensory feedback, relating to the field of emotional intervention technology, and includes the following steps: S1, scene library construction; S2, scene-based preference collection; S3, personalized intervention scene generation; S4, real-time dynamic adjustment. This invention realizes the transformation of intervention content from standardized output to personalized customization. Based on a preference adaptation mechanism of children's autonomous feedback, it can accurately capture each child's unique interests, making the intervention content more aligned with their cognitive habits and emotional needs. This helps to stimulate children's active participation in the intervention process, reducing resistance. Furthermore, it achieves dynamic adaptation between the intervention content and the child's real-time state, enabling timely responses to fluctuations in children's attention and emotional changes during the intervention process. This avoids guidance failure due to rigid content, ensuring the intervention content remains attractive to children, thereby helping to maintain their focused attention.

The application discloses an emotional risk identification method and platform combined with public emotional behavior spectrum, and the method comprises the following steps: extracting behavior features, text features and sound features in historical user interaction sequences, and splicing to obtain multi-source splicing features; based on the multi-source splicing features, outputting public emotional behavior spectrum through a public emotional behavior spectrum identification model and outputting personal emotional behavior spectrum through a personal emotional behavior spectrum identification model; splicing the public emotional behavior spectrum and the personal emotional behavior spectrum for training an emotional risk identification model; obtaining current real-time user complaint data, inputting the emotional risk identification model, and outputting an emotional risk level. By adopting the technical scheme, individual and group emotional behavior spectrum modeling can not only depict the fine-grained emotional evolution path of individuals, but also capture the group emotional aggregation and diffusion process, so that accurate risk grading early warning of personal emotional behavior can be realized in the background of public emotional change trend.

This application provides a method and related equipment for implementing a psychological and emotional companionship service based on an intelligent agent. It relates to the technical field of data processing, including receiving user input information to generate session data, extracting emotional features to construct an emotional feature vector, and constructing an emotional change sequence based on consecutive session rounds to determine emotional trend information; matching the emotional feature vector and emotional trend information with an emotional companionship strategy, generating emotional companionship feedback content and outputting it to the user terminal; performing emotion recognition on subsequent user session data, and adjusting the emotional companionship interaction strategy when the emotion recognition result does not meet preset improvement conditions. By calculating the emotional feature deviation vector and re-matching an alternative strategy when the emotion does not meet preset improvement conditions, dynamic recognition of user emotional change trends and adaptive adjustment of the emotional companionship strategy are achieved, overcoming the problem of accurately recognizing user emotional change trends during emotional companionship interaction.