Persistence in eye movement during visual search

Abstract

As any cognitive task, visual search involves a number of underlying processes that cannot be directly observed and measured. In this way, the movement of the eyes certainly represents the most explicit and closest connection we can get to the inner mechanisms governing this cognitive activity. Here we show that the process of eye movement during visual search, consisting of sequences of fixations intercalated by saccades, exhibits distinctive persistent behaviors. Initially, by focusing on saccadic directions and intersaccadic angles, we disclose that the probability distributions of these measures show a clear preference of participants towards a reading-like mechanism (geometrical persistence), whose features and potential advantages for searching/foraging are discussed. We then perform a Multifractal Detrended Fluctuation Analysis (MF-DFA) over the time series of jump magnitudes in the eye trajectory and find that it exhibits a typical multifractal behavior arising from the sequential combination of saccades and fixations. By inspecting the time series composed of only fixational movements, our results reveal instead a monofractal behavior with a Hurst exponent , which indicates the presence of long-range power-law positive correlations (statistical persistence). We expect that our methodological approach can be adopted as a way to understand persistence and strategy-planning during visual search.

Optimal Control of Eye Movements During Visual Search

“…In contrast to the previous research [10] in our analysis we distinguish between two different types of the multifractality by a calculation of a generalized Hurst exponent for shuffled time series. We separate the time series on fixational and saccadic eye-movements, which allows us to demonstrate the fundamental difference in the temporal structure of these types of eye-movements.…”

“…The time series ∆X and ∆Y were estimated for each trial with certain experimental conditions and concatenated over all participants. After this, we represent the differentiated time series in the following way: ∆X = {F 1 , S 1 , ...., F m−1 , S m−1 , F m }, where F i and S i correspond to the sequences of the movements during time interval of i-th fixation and saccade respectively [10]. We separate the differentiated time series on the fixational and the saccadic time series:…”

“…Saccades are followed by a visual fixation, during which the human oculomotor system generates fixational eye movements involuntarily. Despite the remarkable achievements in the modelling of fixational eye movements and the interpretation of their fundamental properties [1,2,3], there is no comprehensive generic model of fixation selection [4,5,6], which takes into account the underlying mechanisms of visual attention [5,7,8] and qualitatively describes the statistical properties of saccadic eye-movements during the execution of visual tasks [9,10,11,12].…”

PathGAN: Visual Scanpath Prediction with Generative Adversarial Networks

“…When a human observer looks at an image, he spends most of his time looking at specific regions [1,2]. He starts directing his gaze at a specific point and explores the image creating a sequence of fixation points that covers the salient areas of the image.…”

Measurement of Eye Movement in Visual Search for a Target Symbol and Simulation to Construct a Small-World Network

“…Why does eye movement enable us to nd a target symbol swiftly in a visual search? To answer this question, many studies have been conducted on visual search from the psychological viewpoint [5][6][7]. On the other hand, from an engineering viewpoint, we proposed the hypothesis that the eye movements in a visual search construct a small-world network comprising gaze positions (vertices) and steps (paths) on a monitor screen.…”