Research and design of emotional management for the empty-nest elderly

Along with the aging of population, the mental health problems of empty-nest elderly people become increasingly prominent. Speech emotion recognition technology can analyse the speech signals of the elderly and identify their emotional states, thus providing them with timely emotional care and support. This paper focuses on speech emotion recognition and discusses how to use the emotional information in speech signals to achieve emotional monitoring of empty-nest elderly people. This paper first proposes a design of a speech emotion recognition system, then builds an emotional speech database and extracts relevant features of speech signals, and finally, effectively combines the features of speech signals with the existing SVM model, forming the complete speech emotion recognition algorithm, which achieves satisfactory results in recognizing the five basic emotions of the elderly (happiness, sadness, anger, neutral, and fear). The feasibility of the design of the speech emotion recognition system is verified, and a new idea for the development of future welfare technology for the elderly is provided in this paper.