@everyone New docs page for Neon: We just published a new docs page that collects and describes some of our open source tools to work with Neon. These libraries are core building blocks we use internally, and we make them available for researchers and devs to build new applications on top of Neon, integrate with existing workflows, and extend the tools to suit your needs.
Check it out: https://docs.pupil-labs.com/neon/developer/
If you've built something, we would love to see! Please share in 📸 show-and-tell
@everyone Big updates for Pupil Cloud! This one has been in the works for a while, and we’re super excited to finally share it. This update introduces a set of new tools aimed at making eye tracking data easier to analyze, compare, and visualize – all in Pupil Cloud!
Auto Image Mapper (Beta) Auto Image Mapper is a new enrichment that automatically and robustly maps gaze and fixation data onto an image of a planar surface in your scene video. Here are some quick examples of planar surfaces: supermarket shelf, road signs, way-finding signage and maps, restaurant menus, and product packaging. Upload the image you want to map to, and the rest is automatic! We have successfully tested this is many real world conditions.
(We are releasing this as an open beta: Currently free for all to use, but eventually will be part of the Unlimited Analysis Plan.)
AOI Labels With this release, AOIs can be associated with labels. Labels make it possible to classify AOIs flexibly and to aggregate gaze metrics across multiple AOIs and across enrichments within a project.
For example, if you have an AOI of a cereal product on a shelf, it could carry a label "top shelf" and "Cornflakes". This would allow you to calculate gaze metrics for it individually, as well as for all the top-shelf labeled products aggregated.
(part 1/2)
Bar Chart Visualization Use this powerful new tool to visualize aggregate gaze metrics, even across multiple enrichments! Choose the AOI labels on the X-axis and the gaze metric on the Y-axis: fixation duration, time to first fixation, reach, fixation count, and more. Flexibly filter data and build insightful visualizations to include in your presentations.
Privacy feature: Audio Removal on Upload Create a workspace with “Remove Scene Audio” switched on, and audio will be removed from all recordings uploaded to it. This feature is now available for Unlimited Analysis Plans with additional privacy features enabled.
Log in to Pupil Cloud and try out all the new features. Check out the full release notes: https://pupil-labs.com/releases/v8-0-auto-image-mapper-aoi-labels-bar-chart-visualization-and-more
(part 2/2)
@everyone 📰 New Research Digest!
How do expert table tennis players handle rallies that unfold in milliseconds?
Surprisingly, they don’t keep their eye on the ball.
In our latest Research Digest, we highlight a study from the University of Zaragoza and IIS Aragón.
Using Neon wearable eye tracking, they compared professional players, amateur players, and athletes with intellectual disabilities.
Their findings revealed a key difference: While amateurs and athletes with intellectual disabilities primarily tracked the moving ball, experts focused on predictive cues like the opponent’s contact point, using fewer, more efficient fixations to anticipate what happens next.
In one of the fastest interception sports, expertise wasn’t about following the ball. It was about looking ahead!
Read the full Research Digest: https://pupil-labs.com/blog/can-eye-tracking-reveal-visual-strategies-across-table-tennis-expertise-and-cognitive-profiles
Video courtsey of Alejandro Guiseris Santaflorentina (note: see higher resolution video in the research digest post)
@everyone We’ve introduced an update that improves how gaze is calculated, recorded, and streamed across the Neon ecosystem with a focus on flexibility and robustness, while keeping device performance unchanged.
All Gaze Signals, Always Available👀 Binocular, left monocular, and right monocular gaze are now calculated in parallel, saved in every recording, and streamed through both the Real-Time API and Neon-XR. This means every session automatically contains all gaze signals, giving you full access in raw exports and live streams.
In the Companion app, you can select which gaze signal should be treated as the primary source. That selected signal will be used by Pupil Cloud for post-processing, enrichments, analytics, visualizations, and for real-time fixation and saccade detection. You stay in control of which signal drives your Cloud workflows, while still retaining the complete dataset in the background.
Improved RobustnessThis update also strengthens gaze reliability in real-world conditions. If one eye is closed 😉 or partially obstructed, the system handles it more gracefully and automatically falls back to monocular input when needed.
Importantly, Companion device performance remains the same. There is no additional computational load, no change in framerate, and no impact on battery life.
Because this release updates the underlying gaze data structure, we recommend reviewing your integrations to ensure everything continues running smoothly.
To remain fully compatible with the latest Companion App version:
pupil-labs-realtime-api v1.8.0, and you should explicitly select in the app which gaze signal (binocular or monocular) your application uses. Full details are available in the updated developer documentation: https://pupil-labs.github.io/pl-realtime-api/dev/methods/simple/streaming/gaze/NeonXR should be updated to v2.0.0.pupil-labs-neon-recording should be on v2.1.0.PsychoPy, delete the .psychopy3 folder and reinstall the plugin to download the newest dependencies.Neon Player should be updated to v5.0.8 on Windows and Linux. On macOS, you will need to manually set data_format_version to 2.5 in the recording’s info.json file.These updates ensure continued compatibility with the new gaze data structure. If you need additional time to transition, the previous Companion APK is still available here:
If you’d like help reviewing your workflow before updating, we’re happy to support.