DeToX (Developmental Tobii Experiment) is a user-friendly Python wrapper for the tobii-researcher library. It is designed to simplify the integration of Tobii eye-tracking hardware with PsychoPy, particularly for conducting developmental studies involving infants. By streamlining the process of data collection, DeToX aims to enhance the efficiency of developmental eye-tracking experiments.
pip install dvst-detoxor for the latest version:
pip install git+https://github.com/DevStart-Hub/DeToX.gitComing soon to PyPI for even easier installation!
While the official Tobii SDK provides powerful features, we found it could be complex for routine research tasks. DeToX bridges this gap by offering a straightforward approach to eye-tracking data collection.
We don't aim to be the most feature-rich package - we built exactly what we needed for our infant-friendly studies: a simple, well-documented tool that just works
This project didn't start from scratch—it builds upon an existing repository that we have used in the past: psychopy_tobii_infant
While the eye-tracking landscape offers many tools—from PsychoPy's built-in Tobii integration to comprehensive packages like Titta—DeToX carves out its own niche through thoughtful simplicity. We've prioritized clarity and ease-of-use without sacrificing the flexibility researchers need. When your codebase is straightforward and well-documented, it becomes a platform for innovation rather than an obstacle to overcome.
-
Simple Data Recording: Start and stop eye-tracking recordings with just one line of code. Collect gaze data, pupil measurements, and timestamped events automatically during your experiments.
-
HDF5 Data Storage: Save data in analysis-ready HDF5 format with events embedded directly in the gaze timeline for easy analysis, while preserving raw event data in a separate table for advanced processing. Includes comprehensive metadata like screen dimensions, framerate, and recording settings.
-
Infant-Friendly Calibration: Engage infants with animated calibration stimuli featuring zoom and trill effects. Use colorful, attention-getting images with optional sound feedback for better participant engagement.
-
Calibration Management: Save and load calibration files to reuse participant calibrations across multiple sessions.
-
Simulation Mode: Test your experiments using mouse input as a gaze proxy when hardware isn't available. Perfect for development, debugging, and pilot studies without requiring physical eye trackers.
-
Gaze-Contingent Support: Real-time gaze processing for interactive paradigms with configurable smoothing windows. Extract gaze positions using median, mean, or last-sample methods for different latency/accuracy trade-offs.
We believe good software requires great documentation. Explore our comprehensive guides and examples:
📚 DeToX Documentation (coming soon)
Questions or found a bug? Let us know by creating an issue or starting a discussion here on GitHub!