Programmed feeling acknowledgment from the investigation of body development can possibly change computer-generated reality, mechanical autonomy, conduct demonstrating, and biometric character acknowledgment spaces. A PC framework fit for perceiving human feeling from the body can likewise altogether change the manner in which we associate with the PCs. One of the critical difficulties is to recognize feeling explicit highlights from an immense number of descriptors of human body developments. Right now, we present a novel two-layer highlight choice structure for feeling order from an exhaustive rundown of body development highlights. We utilized the component choice structure to precisely perceive five essential feelings: satisfaction, pity, dread, outrage, and unbiased. In the main layer, one of a kind blend of Analysis of Variance (ANOVA) and Multivariate Analysis of Variance (MANOVA) was used to take out insignificant highlights. In the subsequent layer, a parallel chromosome-based hereditary calculation was proposed to choose a component subset from the significant rundown of highlights that expands the feeling acknowledgment rate. Score and rank-level combination was applied to additionally improve the exactness of the framework. The proposed framework was approved on restrictive and open datasets, containing 30 subjects. Diverse activity situations, for example, strolling and sitting activities, just as an activity autonomous case, were considered. In view of the exploratory outcomes, the proposed feeling acknowledgment framework accomplished a high feeling acknowledgment rate beating the entirety of the best in class strategies. The proposed framework accomplished acknowledgment exactness of 90.0% during strolling, 96.0% during sitting, and 86.66% in an activity free situation, exhibiting high precision and power of the created strategy.
CITATION STYLE
Venkata Chalapathi, M. M. (2021). A Novel Two-Layer Feature Selection for Emotion Recognition with Body Movements. In Lecture Notes on Data Engineering and Communications Technologies (Vol. 56, pp. 31–39). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-15-8767-2_3
Mendeley helps you to discover research relevant for your work.