Wyniki wyszukiwania - Biblioteka Nauki

Ten serwis zostanie wyłączony 2025-02-11.

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: jednostka przetwarzania grafiki

Sortuj według:

Ogranicz wyniki do:

Suppressing multiples using an adaptive multichannel filter based on L1-norm

100%

Shi Y. , Jing H. , Zhang W. , Ning D.

Acta Geophysica

2017

tom Vol. 65, no. 4

667--681

Adaptive subtraction is an important link for removing surface-related multiples in the wave equation-based method. In this paper, we propose an adaptive multichannel subtraction method based on the L1-norm. We achieve enhanced compensation for the mismatch between the input seismogram and the predicted multiples in terms of the amplitude, phase, frequency band, and travel time. Unlike the conventional L2-norm, the proposed method does not rely on the assumption that the primary and the multiples are orthogonal, and also takes advantage of the fact that the L1-norm is more robust when dealing with outliers. In addition, we propose a frequency band extension via modulation to reconstruct the high frequencies to compensate for the frequency misalignment. We present a parallel computing scheme to accelerate the subtraction algorithm on graphic processing units (GPUs), which significantly reduces the computational cost. The synthetic and field seismic data tests show that the proposed method effectively suppresses the multiples.

A GPU approach to distance geometry in 1D: an implementation in C/CUDA

80%

Hengeveld S. B. , Mucherino A.

Annals of Computer Science and Information Systems

2022

tom Vol. 30

333--336

We present a GPU implementation in C and CUDA of a matrix-by-vector procedure that is particularly tailored to a special class of distance geometry problems in dimension 1, which we name “paradoxical DGP instances”. This matrix-by-vector reformulation was proposed in previous studies on an optical processor specialized on this kind of computations. Our computational experiments shows that a large speed-up is observed when comparing our GPU implementation against a standard algorithm for distance geometry, called the Branch-and-Prune algorithm. These results confirm that a suitable implementation of the matrix-by-vector procedure in the context of optic computing is very promising. We also remark, however, that the total number of detected solutions grows with the instance size in our implementations, which appears to be an important limitation to the effective implementation of the optical processor.