3D body shape visualisation for online shopping

The European fashion industry generates 13.2% of its turnover via e-commerce. Some time ago, a study indicated that online sales were set to grow by an average of 8.6% annually by 2023. And that was before the coronavirus pandemic reared its head. However, retailers face a number of problems: for example, up to 70% of items ordered online are returned, often because it turns out the garment does not fit.

"In addition, our own survey among 200 people shows that 73% of consumers who want to buy clothing online ultimately decide not to place an order because they're not sure it's the right size," said Geert Mertens, CEO of Shavatar, the newest spin-off of the University of Antwerp and research institute imec. "Of those who do place an order, 40% admit that they tend to order the same item in several different sizes, to make sure they get the right size. These observations obviously have a serious impact on business models in the industry."

3D body shape visualisation

Enter Shavatar, conceived by Femke Danckaers during her PhD research at Vision Lab, an imec research group in the University of Antwerp's Department of Physics.

"I developed a model for predicting and visualising the human body shape in 3D, based on a limited number of parameters, with an average margin of error of just seven millimetres compared to the person's actual body shape," explained Danckaers, who co-founded the spin-off. "Consumers can do this at home, with no need for a scanner."

The technology was developed and refined further in the Shavatar spin-off, with the support of the imec.istart acceleration programme. Geert Mertens said: "We consulted 36 stakeholders from the fashion industry to come up with a tool that would allow consumers to create their own avatars. The tool then suggests the right size and fit of any given item. In a later phase, we will be able to superimpose garments on the consumer's body shape. Our ultimate goal is to be able to visualize entire outfits this way."

In the short term, the tool will be integrated in the online shops of various clothing brands, but if your interest is piqued, you can already visit www.shavatar.me today and create your own 3D avatar. You will receive tailored advice for over 50 brands right away.

The more people use the tool, the more valuable data the fashion industry can gather about consumers' body shapes. These insights will enable clothing brands to further optimize their collections and tailor them to their specific target groups. Since its inception, Shavatar has already raised 200,000 euros in funding, enabling the company to develop a first version of the product.

Imec is pleased with the launch of the spin-off. Jo De Boeck, CSO at imec, said: "Through our spin-offs, we are helping to shape the entrepreneurial landscape in Flanders. We are proud of the incubation process that ultimately resulted in the creation of Shavatar, in collaboration with the researchers and UAntwerp. The success of this collaboration also led to the project’s successful selection for the imec.istart programme."

Professor Silvia Lenaerts, Vice-Rector for Valorisation and Development at UAntwerp, added: "Shavatar is a great example of scientific research leading to applications we can all use. On several occasions now, the Vision Lab research group has managed to translate its extensive imaging expertise into innovative practical solutions. The avatar tool will lead to fewer returns and therefore less transport, which is an important step towards more sustainability.  As an enterprising and sustainable university, we can only applaud that."