Watching a TV show online, a viewer is interested to find out more about a particular dress an actress is wearing. Now, if only she can easily find a shop that sells something similar.
SG:D accredited company Visenze now has such a solution. Its visual tools can seamlessly recommend an online retailer for a similar dress, simply by analysing the video that is playing.
It enables contextual advertisements tied closely to what is showing on screen – and not just the general subject – to be shown to users. Being more relevant to users means the ad has a higher chance of generating revenue.
The magic is in the in-video visual recognition that is performed by Visenze’ technology. Instead of simply relying on text tags, it also analyses the images that are being played and recognises what they are. In doing so, it gives an advertiser an option to deliver a contextual ad to a viewer. Online retailers who use Visenze see a significant increase in conversions, as they are able to surface products that shoppers are already inspired to buy.
In addition, Visenze also has a reverse search capability. This means a user can upload an image and ask a retail shop whether it has a similar item on sale. So, instead of keying in, say, a pair of brown shoes at an online store, a user can share a picture of the shoes. The store, powered by Visenze technology, will recognise the shoes find a match for him.
Visenze originally started as a part of NExT, a leading research centre jointly established between National University of Singapore and Tsinghua University of China.
The company has since partnered well-known names such as Uniqlo, Zalora and Asos to deliver the technology. Because it is based on deep learning, a deeper branch of machine learning, it is dynamic and adapts to changing use cases as well.
Not only is this a smart technology, it is also scalable. With the online experience requiring rapid response from retailers, the key is making the visual technology accurate as well as fast. This means any image can be a potential engagement opportunity.