Semi-supervised training of cell-classifier neural networks

Nowadays, microscopes used in biological research produce a huge amount of image data. Manually processing the images is a very time-consuming and resource-heavy task, so the development and implementation of new automatic systems is required. Moreover, as we have access to a large amount of unlabel...

Full description

Saved in:
Bibliographic Details
Main Authors: Pap Gergely
Grósz Tamás
Tóth László
Corporate Author: Conference of PhD students in computer science (11.) (2018) (Szeged)
Format: Book part
Published: 2018
Series:Conference of PhD Students in Computer Science 11
Kulcsszavak:Számítástechnika, Biológiai kutatás
Online Access:http://acta.bibl.u-szeged.hu/61772
Description
Summary:Nowadays, microscopes used in biological research produce a huge amount of image data. Manually processing the images is a very time-consuming and resource-heavy task, so the development and implementation of new automatic systems is required. Moreover, as we have access to a large amount of unlabeled data, while labels are only available for a small subset, these novel methods should be able to process large amounts of unlabeled data with minimal manual supervision. Here, we apply neural networks to classify cells present in biological images, and show that their accuracy can be improved by using semi-supervised training algorithms.
Physical Description:84-87