Xinyuan Qi ¹ and Garret Washburn ², ¹ USA, ² California State Polytechnic University, USA

As long as online cloth shopping has existed, the issue of seeing how a clothing item looks on a potential customer has always followed. The method proposed in this paper is the FlashFit mobile application and smart-mirror [1]. FlashFit is a system which allows the user to upload images of their own clothing items or ones they've found online and have dynamically resized and fitted onto their image via a recording from the mobile app or in real time with the smart-mirror. The major technologies utilized to create the FlashFit system include Mediapipe for human landmark recognition, the Flutter framework for cross-platform mobile app creation, and a raspberry pi for the smart-mirror [2]. During the development process, the major challenges we faced included the resizing of the clothing item given the position of the user in the camera view as well as the creation of png images from user uploaded clothing item images. Additionally, we employed multiple different experiments within this paper to ensure the FlashFit systems consistency, in which the app performed very well. In result, the FlashFit system is a modern solution to the online clothing market's inability to have fitting rooms and is reliable compared to other solutions currently available.

Virtual Fitting Room, Human Landmark Recognition, Smart Mirror, Online Clothing Retail