BEGIN:VCALENDAR VERSION:2.0 PRODID:-//jEvents 2.0 for Joomla//EN CALSCALE:GREGORIAN METHOD:PUBLISH BEGIN:VEVENT UID:mn5rp9sja7vc6pnqj49u7mupd4@google.com CATEGORIES:RTportal.ru CREATED:20170606T092329 SUMMARY:Machines can see LOCATION:Strelka Institute, Bersenevskaya Haberezhnaya, 14с5А, Moskva, Russia, 11907 2 DESCRIPTION;ENCODING=QUOTED-PRINTABLE:Computer vision is a part of artificial intelligence and is responsible for its ability to recognize images, describe them and extract the most import ant information from the received data. To teach the ‘machine’ in this way to ‘see’ and understand images, events, environment is one of the most diff icult tasks that scientists had to solve in the last decade. The advent of computer vision has had a huge impact on the development of robotics, unman ned vehicles, augmented reality, medical diagnostics and many other industr ies.

The first international summit in Russia and the CIS countrie s ‘Machines Can See’ will enable computer vision specialists to gain access to international experience. At one of the lectures Stanford and MIT profe ssor, author of the book ‘Computer vision: a modern approach’ Jean Ponce wi ll talk about restoring images using computer vision. In addition, INRIA Re search Director in France, Cordelia Schmid, who was awarded the Humboldt Re search Award for her contribution to the study of computer vision, will tal k about recognizing human actions on video clips, and Soumith Chintala, Res earcher at Facebook AI Research, will present overview of the main types of tasks for researchers in machine learning and will talk about modern softw are frameworks for their solution. Among the Russian speakers there will be Victor Lempitsky, Associate Professor at Skoltech, with a lecture on the r ealistic synthesis of the neural image and Dmitry Vetrov, the founder of th e research group of Bayesian methods, with a talk on their application for training neural networks.

Several scientific developments of the p articipants will be selected to be presented to scientists and guests of th e summit. Researchers and development teams will be able to hold a presenta tion in the format of posters or demos. To apply, please, use the summit's website.

During the summit each visitor will be able to check the work of face recognition on himself. For trying this you should upload your photo in advance during the registration. Than the computer with the camer a at the entrance will try to recognize your personality at a glance.
< br/>Summit Machines Can See is held in partnership with Sistema VC and Visi onlabs.

10:30 - 11:00 - Registration

11:00 - 11:15 - Welc ome speech

11:15 - 13:15 - Russian Speakers

Dmitry Vetrov – ‘Deep Neural Networks: Bayesian Perspective’

Viktor Lempitsky – ‘Towards realistic neural image synthesis’

Konstantin Lachman – ‘ Wide nets for all kinds of fish: making multi-task semantic image features work in production’

Alexander Chigorin – ‘Pushing neural networks to production on different platforms. Common problems and possible solution s’

13:15 - 15:00 - Poster and Demo sessions

15:00 - 19:30 - International Speakers

Cordelia Schmid – ‘Structured Models for Human Action Recognition’

Jean Ponce – ‘Machine Learning for Imag e Restoration’

Jiri Matas – ‘Multi-Class Model Fitting by Energy M inimization and Mode-Seeking’

Soumith Chintala – ‘Dynamic Deep Lea rning: New Trends in AI Research’

19:30 - 19:45 - Closing remarks< br/>
http://www.strelka.com/en/events/event/2017/06/09/machines-can-s ee This event was imported from: https://rtportal.ru/index.php/kalendar -sobytij/eventdetail/54/-/machines-can-see?tmpl=component DTSTAMP:20260812T183202Z DTSTART;TZID=Europe/Moscow:20170609T103000 DTEND;TZID=Europe/Moscow:20170609T183000 SEQUENCE:0 TRANSP:OPAQUE END:VEVENT END:VCALENDAR