ShapeMate: A virtual Tape Measure

Abstract

We introduce ShapeMate, a framework for human body shape estimation and classification for on-line fashion applications. Given a single image of a subject our framework is able to simultaneously estimate detailed 3D human body shape and compute foreground segmentation with minimal user input. Once the body shape has been estimated, various semantic parameters are extracted for garment size and style recommendation. Preliminary results demonstrate that a single image holds enough information for accurate shape classification.

Presentation of Historical Clothing Digital Replicas in Motion

“…Related work on human body measurements and body type estimation usually focuses on analyzing the RGB image from which they estimate basic parameters such as body height, hip and waist height [50], or even general body shape divided into several categories like triangle, inverted triangle, rectangle, hourglass, or diamond [51]. Instead of analyzing the RGB image, we chose to use the directly detected skeleton, from which we can calculate the head height, hip height, and shoulder height.…”

Automated Fashion Size Normalization

“…There is currently a variety of work that provides size recommendation. Some tools focus on electronically measuring a user's body shape from users' pictures ( [7], [10]), while others suggest a user's body measurement by using a multiple linear regression approach and a neural network approach when given information on a user's stature, weight, span, and age [5]. However, in a study, authors have found that most of the users who received the correct size recommendation would not buy the size recommended due to fit preferences [14].…”

“…s.t. Constraint (6) The objective function (7) minimizes the weighted pairwise squared difference between normalized sizes across all size types such that the location of the next size must be greater than 0.1 than the previous size in the same size type for all the size types. Constraints (8) specify that all sizes must be greater than 0.…”