Have you found yourself scrolling through Instagram only to see a photo with something you would love to have in your closet? Or to find out more information about where to buy it? We have, and that is why here at Fashwell our vision is to make every image shoppable. Visual search is Fashwell’s leading technology. It was built in an effort to bridge the gap between the content platforms and the purchase path. We want every customer to find exactly what they are looking for. At the same time we want shops to expose to their customers, exactly what they want to see.
The rise of visual search
Our idea that fashion does not fit into a search bar, is at the root of what we do. Search by image is getting customers closer to the checkout path for the exact product they want. It only makes sense that search corresponds to the way in which we find out about fashion, which is image based. With the rise of image search on Google and Pinterest, users are becoming more accustomed to searching with an image.
According to Mary Meeker’s report, 50% of all searches will be images or speech by 2020. Visual search is quickly developing into normal search behavior and soon enough, will come to be expected. For the past 4 years, Fashwell has been working hard at developing a image recognition algorithm so good that any fashion photograph loaded, would return exact or similar products to what appears in the image. According to TechCrunch, picture recognition through deep learning is the biggest leap in artificial intelligence. We are confident that our technology will be one of the best Visual Search providers in the market.
How exactly does visual search work?
Simply put, Visual Search is using a photo for a search where the search results are purely based on the inputted image. Instead of typing text keywords, artificial intelligence analyzes images and detects what is present in the photograph. Fashwell uses computer vision and deep learning, a derivative of machine learning, to power Visual Search. In order to be able to produce results, one needs massive quantities of data. This makes the task not only time consuming but also an arduous task. Many people wonder, does our tech detect color or pattern first, or is it based on shape and size? It is important to understand that deep learning learns from a data set. It does not look at characteristic first but is learning all sorts of characteristics to understand what is in an image.
Fashwell’s Unique Technological Capabilities
Forbe’s tech blog published a good example that has helped some of the business team at Fashwell understand deep learning a bit better:
“The program does this by learning combinations of features that tend to appear together. Cats have visual features – such as their body shape, long whiskers, and the way their faces look – that make them visually different from other animals. The program learns to associate this distinctive combination of features with the word “cat”. This learning process is usually called constructing a model of a cat.” (Forbes, 1 April 2016)
Over the last 4 years Fashwell has been acquiring training data specifically of fashion items to train our algorithms to know the difference between a cross body handbag and a clutch. We have taught it to identify the likes of jumpsuits, dresses, Adidas Superstars and be able to retrieve exact and most similar products to a shop’s products. Currently, our visual search technology is available via an API.
Why choose Fashwell for all your visual search solutions?
Today, it is easy to become overwhelmed by the plethora of machine learning and visual search companies that may show up from a simple Google search. Because our sole focus is fashion, we are constantly tuning our algorithm to deliver the best results. Fashwell’s technology is made accessible within seconds, over an API that easily plugs into front ends and back ends. Our account managers work to find the best solution for your project and will ensure you have a deep understanding of the project scope. We work with each and every company to find the best solution and pricing model that fits perfectly.
Sign up to learn more: