Zobrazeno 1 - 10
of 27
pro vyhledávání: '"Alexander Shekhovtsov"'
Publikováno v:
CVPR Workshops
We consider the training of binary neural networks (BNNs) using the stochastic relaxation approach, which leads to stochastic binary networks (SBNs). We identify that a severe obstacle to training deep SBNs without skip connections is already the ini
Autor:
Alexander Shekhovtsov
Publikováno v:
Lecture Notes in Computer Science ISBN: 9783030926588
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::89ca557a7a01eb74e91c70c36d67f7c6
https://doi.org/10.1007/978-3-030-92659-5_8
https://doi.org/10.1007/978-3-030-92659-5_8
Autor:
Alexander Shekhovtsov, Viktor Yanush
Publikováno v:
Lecture Notes in Computer Science ISBN: 9783030926588
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::60db8b4bfcab8fb57395496a8a83b5d5
https://doi.org/10.1007/978-3-030-92659-5_7
https://doi.org/10.1007/978-3-030-92659-5_7
Autor:
Christian Sormann, Friedrich Fraundorfer, Patrick Knöbelreiter, Alexander Shekhovtsov, Thomas Pock
Publikováno v:
2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
CVPR
CVPR
It has been proposed by many researchers that combining deep neural networks with graphical models can create more efficient and better regularized composite models. The main difficulties in implementing this in practice are associated with a discrep
Publikováno v:
Scopus-Elsevier
In neural networks with binary activations and or binary weights the training by gradient descent is complicated as the model has piecewise constant response. We consider stochastic binary networks, obtained by adding noises in front of activations.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::4422c4bbf022371d659bfc115b144fbb
Autor:
Boris Flach, Alexander Shekhovtsov
Publikováno v:
Computer Vision – ACCV 2018 ISBN: 9783030208899
ACCV (2)
ACCV (2)
In this work we investigate the reasons why Batch Normalization (BN) improves the generalization performance of deep networks. We argue that one major reason, distinguishing it from data-independent normalization methods, is randomness of batch stati
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::83d2693d0661a3479dfdef45d0ec56bf
https://doi.org/10.1007/978-3-030-20890-5_30
https://doi.org/10.1007/978-3-030-20890-5_30
Autor:
Alexander Shekhovtsov
Publikováno v:
Computer Vision and Image Understanding. 143:54-79
We address combinatorial problems that can be formulated as minimization of a partially separable function of discrete variables (energy minimization in graphical models, weighted constraint satisfaction, pseudo-Boolean optimization, 0-1 polynomial p
Publikováno v:
Computer Vision – ECCV 2018 ISBN: 9783030012243
ECCV (4)
ECCV (4)
Dense, discrete Graphical Models with pairwise potentials are a powerful class of models which are employed in state-of-the-art computer vision and bio-imaging applications. This work introduces a new MAP-solver, based on the popular Dual Block-Coord
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::e73c923843769102d401bebeeda0cf8f
https://doi.org/10.1007/978-3-030-01225-0_16
https://doi.org/10.1007/978-3-030-01225-0_16
Publikováno v:
IEEE transactions on pattern analysis and machine intelligence. 40(7)
We consider the NP-hard problem of MAP-inference for undirected discrete graphical models. We propose a polynomial time and practically efficient algorithm for finding a part of its optimal solution. Specifically, our algorithm marks some labels of t
Publikováno v:
2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
CVPR
CVPR
We propose a novel and principled hybrid CNN+CRF model for stereo estimation. Our model allows to exploit the advantages of both, convolutional neural networks (CNNs) and conditional random fields (CRFs) in an unified approach. The CNNs compute expre