Zobrazeno 1 - 10
of 10
pro vyhledávání: '"Uličný, Matej"'
We propose a pipeline for combined multi-class object geolocation and height estimation from street level RGB imagery, which is considered as a single available input data modality. Our solution is formulated via Markov Random Field optimization with
Externí odkaz:
http://arxiv.org/abs/2305.08232
Publikováno v:
IMVIP 2021
Localization of street objects from images has gained a lot of attention in recent years. We propose an approach to improve asset geolocation from street view imagery by enhancing the quality of the metadata associated with the images using Structure
Externí odkaz:
http://arxiv.org/abs/2108.06302
We show how parameter redundancy in Convolutional Neural Network (CNN) filters can be effectively reduced by pruning in spectral domain. Specifically, the representation extracted via Discrete Cosine Transform (DCT) is more conducive for pruning than
Externí odkaz:
http://arxiv.org/abs/2010.12110
Convolutional neural networks (CNNs) learn filters in order to capture local correlation patterns in feature space. We propose to learn these filters as combinations of preset spectral filters defined by the Discrete Cosine Transform (DCT). Our propo
Externí odkaz:
http://arxiv.org/abs/2001.06570
Publikováno v:
European Signal Processing Conference (EUSIPCO) 2019
Convolutional neural networks (CNNs) are very popular nowadays for image processing. CNNs allow one to learn optimal filters in a (mostly) supervised machine learning context. However this typically requires abundant labelled training data to estimat
Externí odkaz:
http://arxiv.org/abs/1905.00135
Convolutional neural networks (CNNs) learn filters in order to capture local correlation patterns in feature space. In contrast, in this paper we propose harmonic blocks that produce features by learning optimal combinations of spectral filters defin
Externí odkaz:
http://arxiv.org/abs/1812.03205
Publikováno v:
In Pattern Recognition September 2022 129
Autor:
Uličný, Matej
Recent discoveries uncovered flaws in machine learning algorithms such as deep neural networks. Deep neural networks seem vulnerable to small amounts of non-random noise, created by exploiting the input to output mapping of the network. Applying this
Externí odkaz:
http://urn.kb.se/resolve?urn=urn:nbn:se:hh:diva-29734
Akademický článek
Tento výsledek nelze pro nepřihlášené uživatele zobrazit.
K zobrazení výsledku je třeba se přihlásit.
K zobrazení výsledku je třeba se přihlásit.
Publikováno v:
Intelligent Computing Systems (9783319304465); 2016, p16-30, 15p