NEW DELHI: As part of its ongoing efforts to use technology in a bid to increase the user experience on its portal, Snapdeal is about to debut image search on its platform. The feature called SnapSearch will allow users to click a picture of something they like, upload it on the website and then browse through similar products available on the platform. Snapdeal’s bigger rival Flipkart also has a similar feature.
Anand Chandrasekaran, chief product officer at Snapdeal told ET that the feature will be live on the website over the next couple of days and will work across several categories. “We are launching it first for watches, women fashion, some categories of men’s fashion etc, and will be expanding it to other categories in the near future,” he added.
Former Bharti Airtel senior executive Chandrasekaran has brought about significant changes to the technology fuelling Snapdeal since he joined the Soft-Bank-backed firm in June this year. The company recently launched a feature called Find My Style, which uses intelligent technology solutions and smart algorithms to pre-empt customer needs. “SnapSearch is one of the iterations of Find My Style,” said Chandrasekaran.
Like Find My Style, SnapSearch too is developed at Snapdeal’s multimedia research lab, which has a commitment of $100 million from the company’s coffers. Chandrasekaran added that what makes this feature unique is its ability to use sophisticated image recognition tools and auto-classify into a category which makes the searches faster.
The product can be used not just by buyers but also by sellers to check similar inventory, proximity-based searches or to check the success of similar items. “There is a lot which can be done through this feature, it all depends on how well it is used,” added Chandrasekaran.
Apart from Flipkart, there are several other companies like Craftsvilla, YepMe and Voonik which have rolled out image search, where the company’s entire catalogue can be searched using images instead of textual inputs provided by the user. To recognise images and show similar results, these companies use computer vision, a growing discipline of computer science where massive computing power is used to give computers the ability to recognise images. Experts consider computer vision to be one of the key building blocks for artificially intelligent machines.
To make this work, large computer networks known as deep neural networks, are trained using a set of images. When the computer is shown images similar to the one it has already seen, it is able to perform a match.
459 total views, 2 views today