OpenET 2 - Remote Eye-Tracking Data Benchmarking Tools

A client-based UNSW capstone project extending OpenET for remote eye-tracking data validation, visualization, and benchmarking.

Problem / goal

Researchers need reliable tooling to find missing samples, timestamp issues, invalid gaze coordinates, incomplete recordings, and inconsistent metadata before running eye-tracking analyses.

My role

Student developer on a client-based UNSW capstone project.

What I personally contributed

• Designed Python data-quality checks for missing samples, duplicate timestamps, irregular sampling intervals, invalid gaze coordinates, incomplete recordings, and inconsistent metadata.
• Built visualization workflows for gaze traces, fixation stability, missing-data timelines, and sampling-frequency plots.
• Contributed user and developer documentation to make the tool easier for researchers to adopt.

Technical approach

• Designed data quality checks for missing samples, duplicate timestamps, irregular sampling intervals, invalid gaze coordinates, incomplete recordings, and inconsistent metadata.
• Developed visualization workflows for gaze traces, fixation stability, missing-data timelines, and sampling-frequency plots.
• Contributed to user and developer documentation for research software workflows.

Key features

• Reusable data validation checks for eye-tracking recordings.
• Quality-control plots for gaze traces and missing data.
• Sampling-frequency visualizations for device benchmarking.
• Research software documentation for users and developers.

Impact / results

• Improved researcher visibility into data quality before analysis.
• Supported maintainable extension of OpenET through documentation and structured Python workflows.

What I learned

• Research software succeeds when validation output is both rigorous and easy for domain users to interpret.
• Client-based projects benefit from small, well-documented tools that fit existing workflows.