Tampere University of Technology

TUTCRIS Research Portal

Cognitive load and metacognitive confidence extraction from pupillary response

Research output: Contribution to journalArticleScientificpeer-review


Original languageEnglish
Pages (from-to)325-334
Number of pages10
JournalCognitive Systems Research
Publication statusPublished - 1 Dec 2018
Publication typeA1 Journal article-refereed


Spontaneous pupillary fluctuations are indicative of the cognitive load imposed while doing a task involving memory resources. However, the fluctuations are also dependent on other factors like lighting conditions, uncertainty or the level of confidence while performing the task and so on. This paper aims to separate various components of pupillary response in order to assess the cognitive load and the confidence with which the task is performed. Hybrid decomposition models using ensemble empirical mode decomposition followed by independent component analysis is found to effectively reconstruct the original signal. The variational Mode Decomposition has been used in order to overcome the limitations imposed by empirical mode decomposition. Results show that variational mode decomposition outperforms existing state-of-the-art methods. Further, we attempted to identify the hidden components of pupillary response during cognitive tasks like mental addition and the anagram test. We obtained Fscore of 0.67 in the detection of cognitive load and Fscore of 0.99 for the detection of confidence level from the single channel pupil data.


  • Cognitive load, Confidence level, EEMD, EMD, Eye tracking, ICA, Pupil size, VMD

Publication forum classification

Field of science, Statistics Finland