Good Buy Clothes Science Popular Online Fitting Room: High Tech Play

After fumbling for more than 2 years, the main attack on the online fitting room has made a breakthrough in technology. It has successfully accessed the online fitting room to the Tmall store of clothing brand, and has attracted two famous VC and ten million dollar B round investments in China and Silicon Valley.

It is learnt that due to technical barriers and high costs, many online technology companies have only stayed at the stage of R & D and trial.

Beginning in 2013, through more than 2 years of exploration, buying clothes has already been able to initially realize users of arbitrary shapes to try out any style clothes of cooperative brands online.

Chai Jinxiang, a co-founder and CTO of the clothing buyer, introduced to the billion state power network that the unique "free ruler measurement" function of the online fitting room for buying clothes is to enable users to input their height and weight and select the characteristics of the body, so that they can get accurate estimation data of thousands of body dimensions.

Based on these data, a good buy clothes will create the exclusive "Medel" of the user (that is, the abbreviation of Me+Model, which refers to the virtual fitting model in person like myself).

Just like the user can see his accurate portrait, even if the shoulder width is different, has the small belly, can display.

Huang Zhongsheng, the founder, describes.

At present, many of the problems in online fitting rooms are mainly in the 3D modeling of clothes. The fitting room needs to show the details of the clothing in three dimensions, and each model needs to be built. This is very time-consuming and energy consuming.

According to the average level of the industry, a skilled worker can only make a model for 1 clothes within 3 days, which determines to a large extent whether the online fitting room can achieve large-scale business.

It is understood that good buy clothes have adopted different technical paths, combining clothing with 3D modeling, avoiding the difficulty of large-scale garment modeling process, and achieving technological breakthroughs.

At present, buying clothes can achieve 3 days for thousands of garments.

After October 2015, good clothes began to cooperate with major clothing brands, including several years of sales of billions of large brands, as well as online sales of hundreds of millions of magnitude of Amoy brands.

Huang Zhongsheng disclosed that according to the pre trial data of the brand store, it was shown that

In the fitting room, the average time for each user to stay in the shop is increased by 3 to 5 minutes, averaging 30 to 60 sets of collocation clothing.

"Although many companies have tried to popularize the online fitting rooms on a large scale, they are subject to technical barriers until the end. We are the only company that can make all the styles of the cooperative brand try on the Internet."

Huang Zhongsheng introduced the current market position of buying clothes.

Huang Zhongsheng said, unable to try on, is currently the clothing electricity supplier, the user can hardly choose the satisfactory style, and the return rate is high, the purchase conversion rate is low one of the important reasons.

This is also a large number of Companies in the world are developing.

The reason.

It is reported that the core technology of buying clothes is mainly from the independent research and development of Dr. Carnegie Mellon and Chai Jinxiang, and the technology team, which combines computer vision, garment engineering, graphics, robotics, image processing and other academic fields.

Well, half of the staff in the clothing buying team are technology. Most of them are Tsinghua, Zhejiang University, Fudan and other famous schools. Huang Zhongsheng himself was a former GGV investor and invested in such companies as beauty, touch and so on.

Chai Jinxiang is a Ph. D. in computer science and a world-class scientist in the field of computer graphics and vision in Carnegie Mellon University.

Taking advantage of technological breakthroughs and efficiency improvements, the market price of online fitting rooms is also very low.

It is reported that the current online fitting room, using 3D modeling method, the modeling cost of each garment up to 2000 yuan to 3000 yuan per piece, and good clothing has been able to achieve a cost of production of only 20 yuan per section.

The brand can only send the corresponding clothing styles and complete the production by buying clothes. After a few days, users can enjoy these styles of online fitting rooms directly in the Tmall store.

Technological origin

It is revealed that a truly commercially available online fitting room technology must meet three conditions: first, the "personalized fitting", so that the users of high, short, fat and thin can see the effect of their personal try; and the two is the authenticity and beauty of the try out, which can not only make the user feel the real details of the clothes, the real situation of the upper body, but also meet the visual requirements of the female; three, "scale", and be able to quickly and low serve the massive clothing styles.

The technical paths adopted by other companies are unable to meet these conditions.

At present, the two most common technology paths in the market are 2D mapping, which is about to synthesize planar photos of clothing into fixed model pictures. Though it is easy to scale, it can not be personalized and authentic.

The other is 3D modeling, which will be modeled on clothing 3D and combined with 3D models with different shapes. Although it can be personalized, it can not achieve scale and authenticity and beauty.