Visual Search – What It Is
More than 60 companies nowadays are working hard to overcome the #1 challenge of the online world — how to search through images and video the same way we search through text. Learn more
Face Detection and Face Recognition
Face detection and face recognition technologies automatically identify a person's face from a digital image or a video source and recognizes its characteristics (like gender, age and nationality). Learn more
Visual Merchandising Monitoring System
Visual Merchandising Monitoring System is a perfect tool to maintain centralized control over goods display in offline stores. Learn more
Duplicates Search and Copyright Infringement Tracking
Duplicates visual search technology detects and tracks copyright protected images in a fast and accurate way. Learn more
Color Image Search
Color search technology lets you navigate your images based on the actual color scheme present. Find images that match one or several colors, and efficiently group them. Learn more
Similar Product Search for E-commerce
Similar4 is visual search technology for online shopping sites. It allows consumers to find similar products by shape, color and other parameters. Learn more
Some of our lovely customers
Face Detection Technology in Odnoklassniki
November 21, Moscow — A famous Russian social network Odnoklassniki (Classmates), project of Mail.Ru Group, announces about a deployment of face detection technology, developed by Kuznech.
Now when all users of a social network Odnoklassniki will search their friends or simply familiar people, they will find more clear content. Firstly they will receive on search request profiles with human faces on avatars, instead of those on which other parts of a body, animals or other foreign subjects are settled.
Methods of algorithm of face detection allow to moderate a content in an automatic mode. The detector is capable to find faces in portrait and group photos, even when there is a head slope on the picture.
Users upload daily from 12 to 18 million photos to social network Odnoklassniki. 7 million photos are personal and go on a moderation. About 10% from them don't pass check.
"One of our main goals is to make the process of searching friends, relatives and acquaintances more simple for our users. We constantly upgrade own mechanisms but to make search querries even more relevant, we agreed about cooperation with Kuznech. We believe that new technologies will change search process of people in Odnoklassniki substantially" — Ilya Grabovsky, the press secretary of a social network told.