Quantitative Measure of Operator’s Experience Based On Gaze Fixation: A Preliminary Study
Year: 2015
Author: Chew, Jouh Yeong; Ohtomi, Koichi; Suzuki, Hiromasa
Institution: Department of Precision Engineering, University of Tokyo
Section: Affect and Experience
Page(s): 146-150
Abstract
An appropriate measure of experience level is rich in information and important for various applications such as physiotherapy and product design. This paper proposes (1) a novel approach to measure experience level based on stochastic model of gaze fixation, and (2) adapted entropy measures to better describe large and sparse model. Content-based Area of Interests (AOIs) analysis is used for gaze fixation to identify attention in dynamic environment. Experiment was conducted by measuring gaze fixation of mobile crane operators for Beginner, Intermediate, and Expert. First, gaze fixation is modeled as Markov chain to summarize transition dynamics between AOIs. Next, characteristics of Markov transition matrices are calculated. Qualitative and quantitative results suggest different response at each level, and the adapted measures exhibit similar pattern as those of transition and stationary entropy proposed by previous study. The model and measures proposed in this study seem appropriate to represent experience level, and further experiment will be conducted to define general AOI settings for different tasks and to verify the preliminary findings.
Keywords: