Pupil Labs Neon

Pupil Labs Neon

Neon is an eye tracking system designed to support scientific research.

Eye tracking for research and beyond
Neon removes the barriers traditionally associated with eye tracking. It’s easy to use: no setup, no calibration. NeonNet, our deep gaze estimation process, delivers unprecedented data quality even in the most challenging environments.

Calibration-free
Neon starts working the moment you put it on. No need to fiddle with nosepieces or camera positions. No calibration required before or during recording.

Any environment
Neon works in any environment and for any activity. From bright, sunny days to pitch black.

Neon = Frame + Module
Neon is all the eye tracking hardware you need in a small package. High-speed eye cameras, a wide-angle camera, stereo microphones, and an IMU, all encased in waterproof silicone. The modular design allows for a wide range of frame shapes, sizes, and styles to fit your research needs.

Prototypists and tinkerers can even integrate the module into their own custom frames.

Frame Types:

Just act natural
An eye tracking device that looks and behaves like normal glasses.
Eye tracking system based on machine learning.
A standard adult glasses frame.
An Android device for real-time eye gaze estimation, user management, recording and real-time streaming.

Better safe than sorry
Professional and efficient.
Eye tracking system based on machine learning.
A standard adult glasses frame.
An Android device for real-time eye gaze estimation, user management, recording and real-time streaming.

Is this thing on?
A lensless, barely visible frame for minimalists and Pupil Core fans.
Eye tracking system based on machine learning.
An adult-sized frame.
An Android device for real-time eye gaze estimation, user management, recording and real-time streaming.

Nothing to see here
Go incognito mode. This frame hides the Neon module.
Eye tracking system based on machine learning.
Standard adult eyeglass frame.
Android device for real-time vision estimation, user management, real-time recording and streaming.

Ready set go!
Eye tracking research solution for sports and other high-dynamic activities.
Eye tracking system based on machine learning.
Adjustable headband shape. Fits well under helmets and other headgear.
Android device for real-time vision estimation, user management, real-time recording and streaming.

I can see clearly now
For users who require vision correction. Frame with a set of interchangeable corrective lenses. Quickly change lenses thanks to a secure magnetic connection.
Eye tracking system based on machine learning.
Standard adult eyeglass frame with corrective lenses from -3 to +3 diopters in 0.5 steps.
Android device for real-time vision estimation, user management, real-time recording and streaming.

Every move you make
A minimal frame for motion capture research.
Eye tracking system based on machine learning.
Adult-sized frame.
Android device for real-time gaze estimation, user management, recording and real-time streaming.
Supplied with 6 x 12mm motion capture markers and mounts.

FOR KIDS

All fun and games
Machine learning-based eye tracking system.
Child-sized eyeglass frames for kids ages 2-8.
Android device for real-time eye gaze estimation, user management, recording, and real-time streaming.

Crawl walk run
Flexible and lightweight frame with integrated head strap for kids ages 2-8.
Machine learning-based eye tracking system.
Android device for real-time eye gaze estimation, user management, recording, and real-time streaming.

Features:
Reduce the lag between data collection and analysis.

Automatically upload recordings to Pupil Cloud.

Your data is securely stored.

Your colleagues can start working with data immediately after it's uploaded.

Use Pupil Cloud to run advanced eye gaze mapping and scene understanding algorithms on your dataset.

Generate qualitative and quantitative insights.

Neon is a video-based, head-mounted eye tracker that uses deep learning to estimate gaze position. Similar to Pupil Invisible, but better. Unlike Pupil Invisible, Neon will provide pupil size and eye state data.

NeonNet was trained using data from multiple people in different environments. The Neon module has two small cameras that capture what the eyes see and one camera that captures your field of view (the scene camera). The Neon module is designed to connect to different frames. The frames have a very simple electrical interface with the module and a USB-C cable that connects to the companion device. The companion device provides power to the module, runs NeonNet, and streams gaze data in real time and syncs with the cloud.

SOFTWARE

Capture: Capture eye tracking data and monitor your experiment in real time with the Neon Companion app.

Analyze: Securely store and analyze your data using advanced gaze mapping and scene understanding algorithms in Pupil Cloud.

Develop: Integrate with tools commonly used in research. Build custom tools using open source libraries.

Get real-time data directly in Neon Companion, including live views of gaze data, scenes, and videos. Easily manage objects and metadata with templates. Play recordings directly on your device and upload to Pupil Cloud.

Recordings can be automatically uploaded to your Pupil Cloud account, so all your data is available in one place. Pupil Cloud lets you share data with colleagues while maintaining full access control. Storage is GDPR compliant and protects privacy.

Offline data
Neon Player lets you easily play and export recordings in convenient formats without uploading through Pupil Cloud.

Build with proprietary APIs and open source libraries: Python, LSL, PsychoPy, Imotions, Unity.

Real-time API
Remote control of all devices: start and stop recordings, annotate events, and more.

Open data format
Save and access all data without any restrictions. Data is saved in fully documented formats. No license restrictions limit your rights to use your data.

Pupil Cloud API
Pupil Cloud API allows you to programmatically access and manage all stored data. You can also run various analysis processes via API to enhance your application.

 

Height 35 mm / Width 40 mm / Depth 10 mm / Weight 7 g
Eye Tracking Technology - Real-time neural network
Eye tracking mode - Binocular, monocular left and right
Calibration - Calibration-free
Slippage Compensation - Slippage invariant. Deep learning powered
Accuracy
- 1.8 deg uncalibrated
- 1.3 deg with offset correction
Scene Camera
- 1600 x 1200 @30 Hz
- H: 103°, V: 77°
- diagonal FOV 128º
Eye Cameras
- 2 x IR eye cameras
- 192 x 192 @200 Hz
IMU - Accelerator, Magnetometer, Gyroscope
Microphone - Dual microphones
Battery Recording Time
- Up to 4 hours
- 25+ hours of recording storage
Cable - USB-C
Computing Device - Mobile computing device: Motorola Edge 40 Pro
Real-time Data
- 2D gaze points in scene camera coordinates at 200 Hz
- Full head pose, accelerometer and gyro data at 110 Hz
- IR eye video showing fused eye regions at 200 Hz
- RGB scene video at 30 Hz
- Pupillometry data and eye state at 200 Hz. Real-time data available in the Companion app.
- Eye position, eye orientation, pupil diameter
Post-hoc Data
- Fixation data: White paper
- Blink data: White paper
Software
- Real-time: Neon Companion App
- Post-hoc: Pupil Cloud
Sample Recordings - Download sample recording
Frame - Multiple frames available.