YouScan Visual Insights

2019: Recognition of logos of brands and objects for a photo in YouScan Visual Insights 2.0

On April 8, 2019 YouScan, the resident of IT of a cluster of Skolkovo Foundation, announced carrying out global updating of YouScan Visual Insights. Logos of brands and objects on a photo taught not only to distinguish a system, but also in detail to analyze a consumption situation: a sex, age, a profession, actions of the depicted characters, colors and even types of a photo and exposure.

YouScan Visual Insights 2.0 is focused on large corporate clients. Daily a system analyzes tens of millions of images online and helps many brands to find useful consumer insights. The added functionality not only expands the existing opportunities, but also open the road for absolutely other cases of application.

The blackamoor Alexey, the CEO and the co-founder of YouScan noted:

Of course, in development of artificial intelligence for the analysis of social networks we should make a lot. But I consider that we already made huge break.

In addition to tracking of visual references on logos, YouScan helps marketing specialists to make the broad picture of an environment of a brand now, distinguishing the following context (data for April, 2019):

• occupations and activities - will be recognized about 60 types of activities on images: run, driving by bicycle, cooking, public statements, swimming, driving of the machine and so forth;
• professions - are defined about 40 names of specialties and a hobby: from musicians and models to weight-lifters and equestrians;
• persons - not only the sex of the person in the photo, but also his age category is defined: baby, child, young man/girl, man/woman, elderly people;
• colors - the color dominating on the image will be recognized that is especially valuable in visual communication with audience: it will become simpler to prove the choice of a color gamma of promotional materials, relying on real digits;
• photo type – selfies automatically separate from a normal portrait photo, monochrome photos from multi-colored, a still life from a portrait, macroshooting from the general photos, commercial photos of a product from UGC (user generated content);
• scenes and events – are analyzed types and types of actions, a situation, time of day, weather conditions and so forth, in total about 430 scenes from consumption situations.

Collecting this information and analyzing it on Big Data, brands will be able to reach the maximum strike accuracy in audience with the advertizing, to make economically justified changes to a product, to find the necessary and most target inflyuenser and also flexibly and to react quickly to a consumer request, offering only what is necessary and as it is necessary.

Anna Gromova, the specialist in researches of the market of the East European region of PepsiCo reported:

For our company it was an opportunity to see how real situations of consumption look. Analytical systems capabilities allowed us to select general options of consumption and also to receive interesting insights at the individual level.