Clothing garment swapping via pose estimation and thin plate splines
Online apparel shopping has huge commercial advantages and convenience contrasted with the traditional shopping but lacks the ability to physically inspect or try on the clothes. To solve this problem, we are working on a virtual try-on technology which has attracted a lot of interest recently by delivering product information similar to that obtained from direct product examination. It allows users to experience themselves wearing diverse garments without efforts of changing them physically. The traditional way of achieving this is by building a 3d model using computer graphics and render the output images. But these approaches require a lot of manual labor or expensive devices to gather necessary information for building 3D models and large computations. In this project, we first perform pose estimation to detect the body joints and these body joints help us identify the corresponding clothing garments. Then, based on the clothing garments, we do body morphing via thin plate splines transformation. We later port the clothing garments from the source to the target human. We conduct different experiments to demonstrate the effectiveness of our work.
Van Tam Nguyen
Primary Advisor's Department
Stander Symposium poster
"Clothing garment swapping via pose estimation and thin plate splines" (2019). Stander Symposium Posters. 1512.