Search Results for author: Pavel Ostyakov

Found 9 papers, 3 papers with code

Recognition of Russian traffic signs in winter conditions. Solutions of the "Ice Vision" competition winners

no code implementations16 Sep 2019 Artem L. Pavlov, Azat Davletshin, Alexey Kharlamov, Maksim S. Koriukin, Artem Vasenin, Pavel Solovev, Pavel Ostyakov, Pavel A. Karpyshev, George V. Ovchinnikov, Ivan V. Oseledets, Dzmitry Tsetserukou

With the advancements of various autonomous car projects aiming to achieve SAE Level 5, real-time detection of traffic signs in real-life scenarios has become a highly relevant problem for the industry.

Adapting Convolutional Neural Networks for Geographical Domain Shift

no code implementations18 Jan 2019 Pavel Ostyakov, Sergey I. Nikolenko

We present the winning solution for the Inclusive Images Competition organized as part of the Conference on Neural Information Processing Systems (NeurIPS 2018) Competition Track.

SEIGAN: Towards Compositional Image Generation by Simultaneously Learning to Segment, Enhance, and Inpaint

no code implementations19 Nov 2018 Pavel Ostyakov, Roman Suvorov, Elizaveta Logacheva, Oleg Khomenko, Sergey I. Nikolenko

We present a novel approach to image manipulation and understanding by simultaneously learning to segment object masks, paste objects to another background image, and remove them from original images.

Generative Adversarial Network Image Generation +1

Label Denoising with Large Ensembles of Heterogeneous Neural Networks

no code implementations12 Sep 2018 Pavel Ostyakov, Elizaveta Logacheva, Roman Suvorov, Vladimir Aliev, Gleb Sterkin, Oleg Khomenko, Sergey I. Nikolenko

Despite recent advances in computer vision based on various convolutional architectures, video understanding remains an important challenge.

Data Augmentation Denoising +4

Cannot find the paper you are looking for? You can Submit a new open access paper.