Results 11 to 20 of about 99,331 (262)

Singular measures and convolution operators [PDF]

open access: yesActa Mathematica Sinica, English Series, 2002
We show that in the study of certain convolution operators, functions can be replaced by measures without changing the size of the constants appearing in weak type (1,1) inequalities.
Aldaz   +8 more
core   +5 more sources

DSConv: Efficient Convolution Operator [PDF]

open access: yes2019 IEEE/CVF International Conference on Computer Vision (ICCV), 2019
Quantization is a popular way of increasing the speed and lowering the memory usage of Convolution Neural Networks (CNNs). When labelled training data is available, network weights and activations have successfully been quantized down to 1-bit. The same cannot be said about the scenario when labelled training data is not available, e.g. when quantizing
Nascimento, Marcelo Gennari Do   +2 more
openaire   +3 more sources

On Convolution Dominated Operators [PDF]

open access: yesIntegral Equations and Operator Theory, 2016
22pages, to appear in Integral Equations and Operator ...
Fendler, Gero, Leinert, Michael
openaire   +3 more sources

Spectral Approximation of Convolution Operators [PDF]

open access: yesSIAM Journal on Scientific Computing, 2018
21 ...
Kuan Xu, Ana F. Loureiro
openaire   +2 more sources

Optimized Separable Convolution: Yet Another Efficient Convolution Operator

open access: yesSSRN Electronic Journal, 2022
عملية الالتفاف هي العنصر الأكثر أهمية في الموجة الأخيرة من أبحاث التعلم العميق. يحتاج الالتفاف ثنائي الأبعاد التقليدي إلى معلمات O(C2K2) لتمثيلها، حيث C هو حجم القناة و K هو حجم النواة. أصبح مقدار المعلمات مكلفًا حقًا بالنظر إلى أن هذه المعلمات زادت بشكل كبير مؤخرًا لتلبية احتياجات التطبيقات الصعبة.
Tao Wei   +4 more
openaire   +3 more sources

Matrix valued p-convolution operators [PDF]

open access: yesOperators and Matrices, 2020
Let \(G\) be a locally compact group with Haar measure, and let \(M_n\) be the set of \(n \times n\) matrices, with the \(C^{\ast}\)-norm, over the complex numbers. Throughout this review \(p\) will denote a real number strictly greater than one. We will write \(L^p(G, M_n)\) to indicate the space of \(p\)-integrable functions from \(G\) into \(M_n ...
Ebadian, Ali, Jabbari, Ali
openaire   +1 more source

FREDHOLM PROPERTY OF COMPOSITE TWO-DIMENSIONAL INTEGRAL OPERATORS WITH HOMOGENEOUS SINGULAR-TYPE KERNELS IN pL SPACE

open access: yesAdvanced Engineering Research, 2014
The authors have previously studied two - dimensional Fredholm integral operators with homogeneous kernels of fiber - singular type. For this class of operators, the symbolic calculus is built using the theory of biloc al operators by V.
Vladimir Mikhaylovich Deundyak   +1 more
doaj   +1 more source

The Dunkl convolution operators and multipoint de la Vallee–Poussin problem

open access: yesVestnik Samarskogo Gosudarstvennogo Tehničeskogo Universiteta. Seriâ: Fiziko-Matematičeskie Nauki, 2013
The Dunkl operator as an object of mathematical physics is considered, we study the kernel and the surjectivity of Dunkl convolution operators in the space of entire functions and the space of entire functions of exponential type.
Karina Raisovna Zabirova   +1 more
doaj   +1 more source

Convolutional Neural Operator

open access: yes, 2023
Repository for the paper Convolutional Neural Operators for robust and accurate learning of ...
openaire   +2 more sources

Image Processing Operators Based on the Gyrator Transform: Generalized Shift, Convolution and Correlation

open access: yesPhotonics, 2019
The gyrator transform (GT) is used for images processing in applications of light propagation. We propose new image processing operators based on the GT, these operators are: Generalized shift, convolution and correlation.
Ronal A. Perez   +2 more
doaj   +1 more source

Home - About - Disclaimer - Privacy