Skip to contents

Elastic consensus matching for encoding and recall fixations

Source: R/gaze_weave_baselines.R
elastic_consensus_align.Rd

This is an inspectable reference implementation of the order-free consensus and moving-least-squares relocation described by Wang, Holmqvist, and Alexa (2021). It is a comparator, not a GazeWeave engine. The iterative relocation follows their Gaussian weighting equations. Their method returns relocated recall and matched encoding fixations rather than a pairwise scalar; `score` is therefore an eyesim-specific, duration-weighted soft match used only as a frozen compatibility comparator.

Usage

elastic_consensus_align(reference, source, spec)

Arguments

reference

Encoding fixation path.

source

Recall fixation path.

spec

A [gaze_baseline_spec()].

Value

An `elastic_consensus_alignment` with relocated points, soft spatial similarity, matching coverage, deformation, and convergence diagnostics.

References

Wang, X., Holmqvist, K., & Alexa, M. (2021). The recorded trajectories of visually guided eye movements can be used to recover the spatial location of previously attended objects. *Attention, Perception, & Psychophysics*, 83, 1700-1716.