Quadratic Recurrent Finite Impulse Response MLP for Indirect Immunofluorescence Image Recognition

Author

Codrescu, Cristinel

Author_Institution

Univ. of Salzburg, Salzburg, Austria

fYear

2014

fDate

24-24 Aug. 2014

Firstpage

49

Lastpage

52

Abstract

The I3A2014 contest participants had to design and implement a pattern recognition system able to classify the cells belonging to HEp-2 images in one of six pattern classes. We propose the QR-FIRMLP architecture, an extension of the finite impulse response multilayer perceptron (FIRMLP). The FIRMLP is a multilayer perceptron where the static weights have been replaced by finite impulse response filters. We have extended this architecture by replacing some sigmoidal layers with quadratic ones and adding recurrent connections to the FIR neurons.

Keywords

FIR filters; image recognition; medical image processing; multilayer perceptrons; FIR neurons; QR-FIRMLP architecture; finite impulse response filters; finite impulse response multilayer perceptron; indirect immunofluorescence image recognition; quadratic recurrent finite impulse response MLP; sigmoidal layers; Biological neural networks; Computer architecture; Finite impulse response filters; Kernel; Neurons; Training; Vectors; HEp-2 cell classification; medical image recognition; neural network; pattern recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Pattern Recognition Techniques for Indirect Immunofluorescence Images (I3A), 2014 1st Workshop on

Conference_Location

Stockholm

Type

conf

DOI

10.1109/I3A.2014.14

Filename

6973548