Visual Search and Foraging

The Virtual Eye project investigates how to predict eye gaze and how to make practical use of these predictive models. Using both mathematical and deep learning models, we aim to create agents that can search efficiently and in a tuneable way.

I have been involved in work on generative models (Bhandari et al., 2025), fixation identification (Orioma et al., 2026), and uncovering new biases with help from step selection methods (Szorkovszky et al., 2025). Ongoing work looks into hybrid modelling, foraging (both empirically and with evolution simulations) and screening for ADHD using reading patterns.

This work is a collaboration with OsloMet AI Lab.

References

2026

  1. Identification of fixations and saccades in eye-tracking data using adaptive threshold-based methods
    Charles Orioma, Josef Krivan, Rujeena Mathema, and 4 more authors
    The European Physical Journal Special Topics, 2026

2025

  1. Modeling eye gaze velocity trajectories using GANs with spectral loss for enhanced fidelity
    Shailendra Bhandari, Pedro Lencastre, Rujeena Mathema, and 3 more authors
    Scientific Reports, 2025
  2. Saccade crossing avoidance as a visual search strategy
    Alex Szorkovszky, Rujeena Mathema, Pedro Lencastre, and 2 more authors
    arXiv preprint arXiv:2508.18404, 2025