The Yangtze River Business News reported that it combines e-commerce data and fashion brand merchandise efficiency data to simplify the problem of wearing
â–¡ reporter Liu Wei
There is always a piece of clothing in the woman's closet. Wearing this great pain point is a huge “uncertain problem†for female consumers. However, with the development of big data technology, the current artificial intelligence is entering the fashion circle to find a solution for this "uncertain problem."
Recently, Mushroom Street revealed that it is “training†a “private wearer†for female users on the platform through machine learning algorithms. However, whether smart matching can capture the hearts of users is still a question.
In an interview with Changjiang Business Daily, the e-commerce expert said that the establishment of AI in Mushroom Street is to build a better understanding of the user, and the relationship between the human brain and the human brain is not large, and it cannot be called true intelligence.
The key to artificial intelligence landing is data
Mushroom Street's “Private Wear Assistant†analyzes and predicts the seemingly intricate trend elements.
How does artificial intelligence help users match?
The relevant person in charge of Mushroom Street told the Changjiang Business Daily reporter that the key to artificial intelligence landing is data. Mushroom Street not only has a large amount of product data, such as product image data, shop red image data, global street data, etc., but also a large amount of user data, such as user-uploaded social image data, video image data, user scene behavior data Wait. Based on large-scale data, Mushroom Street applies machine learning algorithms to train fashion analysis models.
The "little assistant" must first have the ability to understand pictures, that is, image content recognition. For Mushroom Street, letting artificial intelligence understand fashion, we must understand any picture more delicately, and thus identify more detailed elements.
For example, a complete set of full-body pictures is decomposed into "smoke powder round neck letter print T-shirt, blue denim shorts, bright black square bag, white platform shoes". This kind of picture recognition technology can greatly improve the user's search efficiency and provide a solution for matching recommendations.
Zhang Hongming, a senior technical expert in the mushroom street image algorithm who has been engaged in computer vision and machine learning for more than 10 years, told reporters that artificial intelligence and fashion e-commerce combined, but also understand the "combination factors." Matching factors include fashion trends, that is, changes in popular items in different seasons and different regions, in addition to the commodity factors such as fabrics, materials, appearance colors and styles of popular items.
Zhang Xin, director of fashion goods at Mushroom Street, said: “Using big data to insight and predict the trend of commodity trends will not only help the platform to select newer and better style products for consumers, but also to achieve accurate inventory forecasting and promote the supply side of the apparel industry. Upgrade."
"The value of big data lies in the 'processing capability' of data, and the 'value added' of data through 'processing'. The development of data technology can reach consumers' inner needs infinitely, and mastering data is to grasp the needs of consumers." Zhang Xin said.
Whether it is for the traditional e-commerce sector or for fashion brands, data has always been an important measurement dimension.
According to Zhang Xin, in the past, the traditional e-commerce data dimension focused on traffic efficiency, while the data dimension of fashion brands not only paid attention to the sales data in the ordinary storefront, but also focused on the “style of purchase, inventory and salesâ€. effectiveness". The fashion big data of Mushroom Street combines the traffic efficiency data of traditional e-commerce with the commodity efficiency data of fashion brand companies, so that the scope of data coverage is more comprehensive.
Is it true intelligence or a girl?
In 2017, when AI (artificial intelligence) is getting hotter, the application of artificial intelligence in the fashion field has already started. Major companies have entered the market and want to break through the boundaries of fashion and technology.
According to the Massachusetts Institute of Technology, online retail giant Amazon, a research center in San Francisco, has developed a machine learning algorithm that can produce similar styles by learning some specific fashion styles. Put another way, this machine learning algorithm can already be called a simple artificial intelligence fashion stylist or designer.
Recently, the European brand SELECTED also announced the appointment of artificial intelligence Microsoft Xiaobing as its assistant designer, and released the first "skyline series" clothing designed by artificial intelligence. Microsoft Xiao Bing's design process is: by listening to a song, analyzing the song's melody, rhythm, chord color, emotion and other characteristics, combined with the study of different cities' landmark buildings, creating visual works related to the city and song mood.
These are the use of artificial intelligence in the direction of fashion design, but whether these applications can be truly implemented remains to be considered.
Well-known technology e-commerce from the media person Zong Ning said in an interview with the Changjiang Business Daily reporter that "artificial intelligence is currently having a great impact on finance, education and urban management. In fact, it is more important to use the AI ​​fengkou to make new gimmicks. The value is more general at the moment."
E-commerce experts and brand operation experts said in an interview with Changjiang Business Daily that the establishment of AI in Mushroom Street is also to build a better understanding of users, because this is also a change in the development of the times. Today, we talk about AI is only a way for computers to understand and perform human tasks. It has little to do with the human brain and cannot be called true intelligence.
Regarding the development trend of AI, I believe that “the application of AI in the future will be very extensive, such as smart manufacturing, medical health, unmanned warehousing, express logistics, cleaning and sanitation, and even transportation fields such as automobiles, high-speed rails and airplanes. AI, there are many jobs in the future that will replace human beings. It will release huge productivity, save human beings from continuing to work in tedious work, and remind people of the possible unemployment problems, you should go forward."
AI about the impact on the future of human life, Xiao Yu said, "In a sense, some people will be robots 300024, attending stocks ready for him. We have reason to believe that the future in many areas if they can make a simulator Defining the parameters to learn and find the rules by yourself will far exceed the imagination of human beings."
Zong Ning believes that although the challenge of artificial intelligence to human society is obvious, some work will be completely replaced, but at present, worrying about artificial intelligence is a bit of a worry. "Artificial intelligence does not require sensory stimulation or psychological satisfaction, so I think it is quite unreasonable to think about the machine entirely by human thinking."
Editor: ZB
Metal Belt buckle, men's belt buckle, fashion buckle, new style buckle, high belt buckle, Material is brass, stainless steel , zinc alloy, etc.Technics is Carved, casting, painted (nickel,brass,gold,chrome,rhodium) polished, welding, etc, Multi-color is available, It`s widely used for Gifts, crafts, or collections, and souvenirs, decorations, promotionals, etc.
We produce all kinds of buckles, which possess good quality, customer's design are welcome. We can mark any logo or trademark on the buckle.
Belt Buckles For Men,Brass Buckle,Metal Buckle,Belt Buckle
Shenzhen MingFengXing Art & Craft Products CO., LTD. , https://www.customizedbadge.com