Pupil Cloud - Reference Image Mapper
Author(s): Pupil Dev Team
April 8, 2021
Out of the box Pupil Invisible can provide you with qualitative insight into where the participant is looking.
For quantitative analysis you need to know what participants are looking at as they move through the world. Traditionally answering the what question, required many hours of human coding to manually map gaze from each recording onto a single image of the object of interest.
Reference Image Mapper
Our new enrichment - Reference Image Mapper - makes it easy to compare multiple subjects’ gaze relative to the same real-world object. We use state-of-the-art computer vision algorithms to track the objects visible in a reference image and then automatically map gaze from your recordings onto the image. No manual coding or markers needed!
You can easily obtain a heatmap of aggregate gaze data overlaid on the reference image or download the mapped gaze data in CSV format for further processing. Please note that this enrichment is a beta feature. Let us know if you have any feedback!
The Gaze Overlay enrichment allows you to download world videos with a gaze overlay rendering. You can easily filter out the sections of your recordings you are interested in and the visualization of the gaze circle can be customized (currently: size and color).
Raw Data Exporter
Use the Raw Data Exporter to get access to all your recording data in convenient CSV and MP4 format. Easily export data from an entire project and continue your analysis with your tools of choice (e.g. R, Python, third party platform). In addition to the raw recording data it also includes event annotations you have added post-hoc.