One of the challenges of marketplaces or e-commerce with thousands or millions of products is the image tagging task.
The growth in the number of products being sold online and the relative heterogeneity in the categories of these products has become physically impossible and infeasible to manually tag them. Product tagging is a must to simplify navigation and deliver accurate results that convert.
Some product catalogues have thousands of different brands, and a global customer base, meaning an intelligent taxonomy is important for creating “credibility at scale”. In fashion, there are so many ways of describing the same thing, as there are as many tagging criteria as perceptions.
In other words, not everyone will tag the same images using the same words. This leads to discrepancy in the kinds of tags assigned to the products which minimizes the capacity of searchability of products. And meeting the need of users to be accurate in taking them to the right product is essential.
For example, what I call a “duvet coat” might be a “down jacket” to you and a “puffer coat” for Zara.
Customers use many different words to describe the same piece, so comprehensive, specific attributes help recommendation engines make associations between related garments. Using auto-tagging exclusively provides several benefits over manual tagging: saving you a lot of time and effort!
In addition to improving fashion search results, AI-enhanced taxonomy can better identify products on a spectrum, rather than in binary terms. For example, a human might say an item either is or isn’t a “cold shoulder” top, while a machine can identify it as 80 per cent “cold shoulder”, due to smaller sleeve openings.
An automatic image tagging system like Wide Eyes’ Auto-tagging API helps taking care of both of these problems and builds an efficient product tagging system even if the database consists solely of visual information about the products. Wide Eyes’ automated solution will set homogenous tags around the whole catalogue. For these reasons the smart organizations are using AI-powered visual recognition technology to automatically extract product attributes from fashion images.
Conclusion…
Automating the product tagging process, retailers save time and costs, and improve efficiency in catalog management – from product upload and right categorization to SEO and product ranking.
Auto-tagging based on visual features is helping to index the products better leading to more accurate searches on the site. The more relevant tags a product image has, the greater probability it has to appear in specific search results from shoppers.
Wide Eyes’ Auto-tagging API is easy to implement, in less than 24 hours it could be in full production. If you want to know how to set up our auto-tagging API, contact us!
