The Third Christmas Colloquium on Computer Vision

Tuesday, 26 December 2017, full day starting at 10:20

Center for Data-Intensive Science and Engineering (CDISE) of Skoltech invites you to the colloquium, where nine researchers will present their state-of-the-art work related to computer vision. Each talk will have an introductory part to make it accessible to a wider audience. Some minimal level of familiarity with modern computer vision is assumed. The talks will be in English.

Attendance is free of charge, but registration is required, please fill in this form.

Venue: Skolkovo, “Technopark” office center, Mendeleev Quarter, Building 3 (Mendeleev quarter consists of four buildings, Skoltech is in the dark blue one).

Getting by the special bus: the bus from the “Park Pobedy” metro will depart at 9:40. SEE MAP  FOR THE BUS STOP LOCATION.
Getting by car
Getting by public transport: Take any bus that goes to Skolkovo/Parkovka Minskaya or Skolkovo/Kvartal Mendeleeva. Kvartal Mendeleeva is the stop you need, the colloquium is in the Dark Blue Building.


9:40 — The bus departs from “Park Pobedy”

Session 1: Computer Vision In the Wild
10:20 — 11:00  Alexander Vakhitov (Skoltech): Set2Model Networks: Learning Discriminatively To Learn Generative Visual Models  [slides] [youtube]
11:00 — 11:40  Natalia Neverova (Facebook AI Research): Multi-Person Human Pose Estimation In The Wild [slides] [youtube]

11:40 — 12:00   Coffee break

Session 2: Efficient Convolutional Networks
12:00 — 12:40  Sergey Zagoruyko (ENPC Paris): Weight Parameterizations in Deep Neural Networks  [slides] [youtube]
12:40 — 13:20  Michael Figurnov (Higher School of Economics): Spatially Adaptive Computation Time for Residual Networks [slides] [youtube]

13:20 — 14:20   Lunch

Session 3: Beyond Computer Vision
14:20 — 15:00  Sergey Bartunov (DeepMind): Fast Adaptation in Generative Models with Generative Matching Networks [youtube]
15:00 — 16:00  Alexey Dosovitskiy (Intel Labs): Issues with Deep Reinforcement Learning [slides] [youtube]
16:00 — 16:20  Coffee break 

Session 4: Generating Images
16:20 — 17:00 Alexander Kolesnikov (IST Austria): PixelCNN models with Auxiliary Variables for Natural Image Modeling [slides] [youtube]
17:00 — 17:40  Anton Osokin (Higher School of Economics): GANs for Biological Image Synthesis [slides] [youtube]
17:40 — 18:20  Dmitry Ulyanov (Yandex/Skoltech): Deep Image Prior [youtube]

18:35 — The bus departs to “Park Pobedy”

“Emergency” phone: +7 916 391 47 73 (Victor Lempitsky)





 melnichek                         mapbox-logo-color-1

The meeting is also supported by the Ministry of Education and Science grant RFMEFI61516X0003.
The 2nd Colloquium webpage can be found here.

The 1st Colloquium webpage can be found here.

See you at Skoltech!