Research on the Technology of Searching for Fashion Trend Image Based on ResNet50 Model

With the further development of technology, providing technical support for the forecast and development of fashion design trends through artificial intelligence has gradually become a reality. In order to enable users to obtain more accurate design trend materials in a shorter time, this paper presents a 'searching by image' model of clothing trends based on the ResNet50 model. Through the targeted collection of data and the rational construction of algorithms, the system can retrieve and output more image materials with relevant design features for the existing image materials for the development of design trends as well as the development and design of styles. By comparing the system with Google's and Taobao's 'searching by image' system, it is concluded that this system has high efficiency, high accuracy and high correlation.