Autonomous analysis of confocal images: using machine learning to recognize bijels

Condensed Matter lunchtime seminar

Autonomous analysis of confocal images: using machine learning to recognize bijels

  • Event time: 1:00pm until 2:00pm
  • Event date: 18th March 2019
  • Speaker: (School of Physics & Astronomy, University of Edinburgh)
  • Location: Room 2511,

Event details

We are developing a protocol using machine-learning techniques to allow the identification of a bijel from a single image. We use confocal data and classification judgements from previous bijel experiments, and process them in order to obtain parameters that may be useful for bijel identification, such as features of the autocorrelation function. Our approach is also capable of including composition information from the experiments into the classification process.
 
We then use some of these parameters to categorise images into bijel and not-bijel, using machine-learning algorithms. The algorithms allow us to identify which parameters are key for identifying the bijel, as well as generating a model that can be used to quickly and easily categorise future images from new samples. We compare several machine learning algorithms as well as an array of possible parameters, in order to achieve the lowest error rate possible when classifying bijels.

About Condensed Matter lunchtime seminars

This is a weekly series of informal talks given primarily by members of the soft condensed matter and statistical mechanics groups, but is also open to members of other groups and external visitors. The aim of the series is to promote discussion and learning of various topics at a level suitable to the broad background of the group. Everyone is welcome to attend..

Find out more about Condensed Matter lunchtime seminars.