Visual Journeys: Eye Tracking and Abstract Art Perception

Eye tracking offers valuable insights into visual perception, making it useful in understanding how we engage with art. In their study, Modeling Perceptual Organization in Abstract Art Using Eye Tracking Data, Carlo Ram M. Ferrer and Maria Mercedes T. Rodrigo explored how individuals with varying levels of art expertise perceive abstract art. Their research investigated the Gestalt principle of proximity, which suggests that closely grouped elements are perceived as related. Here’s how RealEye contributed to their findings and what the data revealed.

RealEye: A Window into Perception

Ferrer and Rodrigo used RealEye to record gaze data from 20 participants, aged 18 to 32, categorized into novice and expert groups. Each participant viewed six abstract art pieces for 30 seconds per stimulus, with gaze data captured in real-time. The eye-tracking experiment was conducted remotely via Zoom, using participants’ webcams to record gaze data.

Six digital abstract art prints were carefully selected in consultation with art experts. These stimuli demonstrated Gestalt principles, particularly proximity, to explore how participants grouped visual elements. Participants viewed each art piece with the sequence randomized to minimize order effects.

Eye Tracking Metrics and Analysis

The study employed a mix of RealEye’s built-in capabilities and additional analytical methods to explore participants’ visual perception of abstract art.

Metrics Captured Using RealEye

1. Fixation Points and Durations: RealEye recorded fixation points - the locations where participants’ gazes lingered - and their durations. These fixations indicated areas of high visual interest and helped identify participants' focus within each abstract art piece.
2. Raw Gaze Data: RealEye exported raw gaze data at intervals of 30ms, capturing precise eye movements throughout the 30-second viewing window for each stimulus. This dataset included x and y coordinates for each gaze point, which formed the basis for identifying scanpaths and clustering.
3. AOI-Based Exports: Areas of Interest (AOIs) were computationally defined based on clustered fixation data. RealEye’s export provided an overview of fixation density in these AOIs, facilitating an understanding of which parts of the art drew the most attention.

‍‍

Analyses Beyond RealEye

While RealEye collected the foundational data, additional analysis tools were employed to extract further insights:

1. Scanpath Pairwise Distance Analysis: The researchers used Eyenalysis, a Python-based tool, to compare participants’ scanpaths (sequences of eye movements). This analysis calculated the spatial and temporal similarities between scanpaths within and across the novice and expert groups.
2. Gaze Clustering with OPTICS: To analyze gaze patterns, the OPTICS (Ordering Points to Identify the Clustering Structure) algorithm was applied. This machine learning technique identified high-density gaze clusters, indicating regions of high visual engagement. The researchers then calculated Intersection over Union (IoU) scores to measure the alignment of gaze clusters between participants and the expected AOIs.

What Our Eyes Reveal: Key Findings

The study revealed intriguing patterns in how participants perceived abstract art. Despite the Gestalt principle of proximity suggesting that closely grouped elements should be perceived as related, both novice and expert groups showed low adherence to this rule. IoU scores, which measured the alignment between participants’ gaze clusters and expected areas of interest (AOIs), averaged below 0.35. This fell short of even a “decent” similarity threshold (IoU > 0.5). Interestingly, higher IoU scores were observed during the first few seconds of viewing, indicating stronger initial alignment that faded over time.

Statistical analysis using Welch’s t-Test showed no significant differences between the gaze behaviors of novices and experts. Both groups exhibited similar scanpath patterns and clustering behavior, suggesting that formal art training had minimal influence on their ability to visually organize elements according to the principle of proximity.

These findings emphasize the subjective nature of art perception and suggest that the Gestalt principle of proximity may have limited applicability in the context of abstract art.

Conclusion

Ferrer and Rodrigo’s 2024 study highlights the value of RealEye in academic research and its potential to deepen our understanding of visual perception in abstract art. While the results challenge the universal application of Gestalt principles, they pave the way for further exploration of gaze behavior and perceptual organization across various contexts. Future studies could benefit from refining clustering methods and expanding the scope to other types of stimuli.

You can run a similar study!

Follow the steps below to start your own experiment with RealEye:

1. Go to RealEye Dashboard and create or log in to your account.
2. Purchase the License of your choice (https://www.realeye.io/pricing). If you need any custom adjustments, contact us at contact@realeye.io. We are happy to help!
3. Activate your license by following the instructions in the RealEye License Activation Guide

‍

Ready to set up your own study? Visit RealEye Support page to learn more and keep us posted on your results! 🚀