Researchers increasingly are using machine learning to eliminate noise from microscopy images.
Machine learning involves teaching the computer how to de-noise one set of images and then apply what it has learned to new data.
Supervised models that are trained with matching pairs of noisy and clean images improve as more input pairs are provided, but the lack of pre-cleaned images has spurred the development of algorithms that train themselves.
Meanwhile, tools are being developed that enable researchers to compare multiple de-noising approaches and contribute new ones.
These include CSBDeep, an online machine-learning toolbox developed by Germany's Max Planck Institute of Molecular Cell Biology and Genetics computer scientist Florian Jug, which can be used with the Fiji image-processing environment or Python programming language, and ImJoy, a one-stop-shop for test-driving multiple machine-learning methods developed by Sweden's KTH Royal Institute of Technology computer scientist Wei Ouyang.
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA