Application of costume design feedback system based on IoT platform and nonlinear computing
Abstract
With the proliferation of artificial intelligence and cloud computing, the integration of affective computing and intelligent cloud platform technology has emerged as a nascent trend within the realm of fashion. This research endeavor, aligned with the principles of Nonlinear Engineering-Modeling and Application, seeks to delve into the realm of personalized apparel design by elucidating the utilization of affective computing to discern and address customers' emotional requisites. The present study commences by establishing a cloud-based platform for data collection, harnessing the power of image information alongside electronic questionnaires facilitated by the Internet of Things (IoT) technology. Subsequently, a facial expression recognition system is erected, employing Gaussian label construction, an enhanced separable convolutional neural network (CNN), and an attention mechanism, showcasing the application of Nonlinear Engineering-Modeling and Application principles. In amalgamation with questionnaire information, the aforementioned system demonstrates an average recognition accuracy of 90.7% across three distinct emotional states, surpassing the performance of conventional CNN approaches. Ultimately, an analysis of consumer satisfaction pertaining to this product batch is accomplished, hinging upon the recognition data, thereby fostering innovative concepts for future user feedback encompassing clothing design and the domain of affective computing. The findings of this study contribute to the field of Nonlinear Engineering-Modeling and Application, providing valuable insights and methodologies for the development and implementation of intelligent fashion systems that align with the principles of Nonlinear Engineering-Modeling and Application.