“A picture is worth a thousand words”- There is scientific evidence that 55% of communication is visual imagery, 38% is your tone and only 7% account for words. Complex ideas can be conveyed by a single still image, which conveys its meaning or essence more effectively than a mere verbal description. Ever played Chinese whisper? In this game, the right message has never reached the final person in the line. What if we replace the whisper with an image?
For any system, the quality of output is always proportional to the input received. Rather than translating what you see into words, sending the image directly is the best possible input. The best possible input is to take a picture and let the system figure it out. It also saves time to type it out. Better description means more time spent in typing. In this era of 160 characters, it’s not desirable.
For an AI-driven system, an image is an input without a user bias. Consider different people searching for the same piece of jewellery advertised in a magazine. The description of jewellery would vary according to the vocabulary level of the users. This variance is considerably less when the users input an image.
Smartphones with cameras are ubiquitous. A piece of technology enabling an input to an AI system where you can give the input as clear as possible. The system can identify the object(s) of interest and show you an output that segregates the image received into the items that are part of the image. These items present would include your item of interest. Now if you want, the system has a capability to find out items that are exact or similar to the item recognized. Also, if present in an online store’s catalog it can be purchased.
For businesses having an e-com site or a mobile app, the visual search can be integrated as an add-on to enable the end-users to search for items that they are seeing and might want to purchase. It can range from what they see in a store, on the sidewalk, in a magazine, on a ramp, in a movie… The list is endless. One such solution for retailers developed for improving the user experience of end users available at: https://appsource.microsoft.com/en-us/product/web-apps/mozantatechnologiesllc.mozcom_ai_001?src=office&tab=Overview
One specific advantage of the solution is the ability for the retailer to control the output received by the end users from the AI.
