Page 26 - Anthropometry, Apparel Sizing and Design

New directions in the field of anthropometry, sizing and clothing fit 21

0
reality and augmented reality fit programs allow the customer to experience a 360
retail experience within the privacy of their home, allowing them to make data-
supported intelligent clothing choices. Some of the platforms are data-driven, while
others are strong in visualization effects and tech-intensive depending on the compo-
sition of the team Gill (2015). These interfaces can be classified into those that offer fit
recommendation (Fitbay, Fits.me) or size recommendation (Dressipi and True fit) or
fit visualization (Metail, My Virtual Model) or a combination of all three. These vary
widely in how they communicate with the consumer and how they convey fit and
drape (Miell, 2018).
The basis of all virtual trial rooms are the 3-D avatars. The avatar can be a generic
body or a scan of the customer or a parametric avatar that can be modified to simulate
the body of the customer. Animated avatars that simulate postures can also be pro-
vided for a more realistic simulation of garment fit during actual use. Most of the cur-
rently available avatars are rigid and do not realistically portray the deformations of a
soft body when a garment is worn, for example, the compression of the soft breast
tissue by a bra or bikini top. In a recent development, Harrison et al. (2018) reported
the development of soft body avatars that simulate the compressibility of body tissues.
Any rigid 3-D scan can be converted into a soft avatar. Using finite-element methods
the behavior of a body-fitting garment on a deformable body can be simulated in a
realistic manner. Stresses and strains in the garment and the body are simulated
together, with two-way coupling of forces and displacements.
A number of companies are now proposing solutions to size and fit issues Gill
(2015). These solutions aim to engage the customer in the garment purchase process.
The customer is asked to input details and some body measurements, the system
makes some recommendations, and the customer can visualize the options in a virtual
trial room before making the final purchase decision. Initial results from these systems
indicate that several consumer groups such as the senior women buyers have low fash-
ion confidence. For such customers, visualization of fit alone is not enough. They pre-
fer the fit and size recommendation tools that offer expert styling advice on what
would look good on them or which current fashion trends would suit them
(Miell, 2018).
Data scientists and styling experts are working together to build highly intelligent
recommendation tools to engage and guide the customer to make the right choices.
Customers input height, weight, and bra size data into the system (Limited, 2019),
which are used to compute waist and hip dimensions. The tool makes a size recom-
mendation based on these inputs. Recommendations regarding what style to buy are
made by algorithms based on the details provided by the customer regarding their
body shape, coloring, life style, age, and previous buys and returns (Dressipi.com,
2019). Data analytics are used to analyze what is returned most and relate it to specific
body shapes and sizes, and algorithms are modified so as not to make those recom-
mendations (Miell, 2018).
Some systems take the consumer experience further by making recommendations
about the total look. For example, while trying on a party jacket in the virtual trial
room, recommendations are made regarding the bags and shoes that will go with
the outfit. These recommendations are based on previous choices made by customers