Militants launch shell on exhibition complex near Damascus - televisionWorld August 20, 15:27
Cardinal Parolin: Dialogue of Roman Catholic and Orthodox Churches to help them feel unitySociety & Culture August 20, 8:27
Polina Dibrova, mother of three, wins Mrs. Russia 2017 beauty pageantSociety & Culture August 20, 4:41
Russian emergencies ministry plane returns from firefighting mission in ArmeniaWorld August 20, 4:39
East Ukraine conflict claimed nearly 3,000 civilian lives — ICRCWorld August 20, 1:56
Renowned Russian filmmaker Andrei Konchalovsky turns 80Society & Culture August 20, 0:48
One of seven injured in Surgut stabbing spree in critical condition — authoritiesSociety & Culture August 19, 23:51
Netanyahu expects to meet with Putin in Sochi on August 23 — Israeli premier’s officeRussian Politics & Diplomacy August 19, 22:47
Surgut attacker is identified as a local resident - investigationSociety & Culture August 19, 14:09
MOSCOW, September 12. /TASS/ Skoltech and Inspector Cloud company have recently initiated the development of a system capable of recognizing images to account for goods on the shelves of grocery stores, the Skoltech press-service reported.
Up to now, accounting for wares in Russia has been performed manually, therefore, the data on the actual presence of an item and its amount on the shelves might be unprecise.
Inspector Cloud receives pictures of grocery shelves stocked with consumer items and with the algorithms of its neuron networks recognizes the goods and then generates a report of the actual status. Skoltech scientists will consult the developers from Inspector Cloud, and share their experience and knowledge in the field of computer vision.
From Skoltech’s side, Viktor Lempitsky will be taking part in the joint project. In the past, he was employed at Microsoft Research and at the Massachusets Technology Institute, now he leads a group of computer vision gurus at Skoltech. He is one the most prominent researchers in the field of technical vision and neuron networks.
The tasks of automatic recognition are employed in others fields, for instance, in identifying a person by a photo or a video or for other types of clouds. -0- /pa.