英语翻译In this paper,we propose a flexible method that,in princi
英语翻译
In this paper,we propose a flexible method that,
in principle,can be used for super-resolution problems
with arbitrary magnification factors up to some fun-
damental limits.More importantly,we propose a new,
more general way of using the training examples,so
that multiple training examples can contribute simulta-
neously to the generation of each image patch in the
high-resolution image.This property is very important
as generalization over the training examples is possible
and hence fewer training examples are required.
The rest of this paper is organized as follows.In
Section 2,we formulate the super-resolution problem
more precisely and introduce our method based on ideas
from manifold learning.Some details of the experimen-
tal setup are discussed in Section 3,including feature
representation,training set,and model parameters.Ex-
perimental results are then presented in Section 4.Fi-
nally,Section 5 gives some concluding remarks.