Answers about the wearable binocular eye tracker from Argus Science: what comes with it, how gaze is measured, what data it records, and where it can be used.
Manufacturer's information · Last updated 15 October 2026
The full figures are in the ETVision specification in the Download Center. Questions about live object recognition are answered on the AI LMOT page.
The ETVision optics and controller, three nose pieces for different head and nose sizes, a removable sun shield, a calibration target, a carry case, a 9 V power supply with a controller belt, and a pre-formatted 128 GB microSD card. The software comes with it: the ETVision application, AI LMOT with a standard 80-object model, one ETAnalysis license, ETRemote, SDK samples with source code, and ETPhone for iOS or Android. Updates are included, and there is no annual license or maintenance fee.
ET3Space places gaze in room coordinates on up to 20 real surfaces and objects; it needs a motion-capture system. AIObjectTraining trains AI models on your own objects. StimTrac analyzes screen-based studies recorded with the wearable. The Stationary Scene Camera adds a synchronized room-mounted view. EEG TTL Sync provides a precise pulse for EEG and other instruments. Additional ETAnalysis licenses are available for more analysts or workstations.
More than four hours on one battery charge. Storage is not the limit: the included 128 GB microSD card holds well over four hours of recording, the controller accepts cards up to 400 GB, and it can also run from its 9 V supply while recording. The recording goes to the card in the controller, so it does not depend on a laptop or a phone staying connected.
Yes. The optics weigh 56 g and sit over whatever glasses the participant normally wears, single-vision, progressive or bifocal, so they see the task as they always do. The controller rides on a belt or in a pocket, and the only cable runs from the glasses to it. Recording continues without a computer nearby, outdoor use is supported, and ETPhone lets the investigator calibrate and control recording from a phone. ETVision has recorded running, cycling, mountain biking, alpine skiing, baseball, football, lacrosse, golf and driving, and Argus Science works with professional football and baseball teams. Examples are on the Argus Science YouTube channel.
Yes. Both eyes are measured at 180 Hz, one sample every 5.56 ms, while the recording runs, and the rate holds with AI LMOT running. Nothing has to be reprocessed afterwards to reach it. Both eye videos are recorded at the same 180 frames per second, so the rate and every sample can be checked later.
No. ETVision measures the dark pupil and the corneal reflection in each eye image and computes gaze geometrically. Vergence and the 3D gaze point are solved from both eyes as geometric computation, not estimated by a model. The same inputs always give the same answer, and any result can be checked against the recorded eye video. AI is used in ETVision for recognizing objects in the scene, not for computing gaze.
Yes. Each eye is measured independently, and ETVision reports the 3D line and point of gaze relative to the head, solved from binocular vergence geometry. With ET3Space, gaze is reported in room coordinates on named surfaces.
Both eye videos are recorded at 180 frames per second, so the session can be replayed frame by frame and recalibrated afterwards. A session that would otherwise be lost can usually be recovered instead of recollected. Slippage compensation and parallax compensation are both built in.
At 180 Hz: point of gaze on the scene image, the 3D line and point of gaze, fixations, blinks, pupil diameter for each eye in millimeters or pixels, eyelid position, corneal reflection position, and gaze on AI LMOT objects or live areas of interest. Both eye videos are recorded at 320 × 240 and 180 frames per second, and the scene video at 1080p and 30 frames per second, with sound. Data records to .csv or binary files and streams over the network.
ETAnalysis, which is included, reduces gaze to fixations, saccades and dwells, tracks moving areas of interest, and produces heat maps, scan paths, swarm displays and group statistics, with export to Excel or text. ETVision also works with iMotions and BIOPAC AcqKnowledge for analysis, with BIOPAC Systems and ANT Neuro for physiology, and with NDI, Advanced Realtime Tracking, OptiTrack, Vicon, Qualisys and other motion-capture systems through ET3Space.
The EEG TTL Sync option sends a precise TTL pulse to EEG or other equipment, and external systems can send 16-bit event values over the network, recorded with the gaze data. Live data and video stream over Wi-Fi or Gigabit Ethernet with ETRemote, which also controls recording and calibration remotely. SDK samples with source code are included in C++, C#, Python, MATLAB, Lab Streaming Layer and MFC.
Yes, with the StimTrac add-on for ETAnalysis. It maps head-mounted gaze onto a stationary screen, as if the data came from a desk-mounted tracker, and areas of interest defined once apply to every participant. One wearable can then cover both kinds of study.
Yes. Gaze computation, object recognition and analysis run on your own system, with no connectivity and no account required, so classified, air-gapped and regulated sites use the same workflow as everyone else. ETVision is designed, built and supported in Massachusetts by Argus Science, a US-owned company that is ITAR compliant.
The eye cameras use non-coherent LED illumination at 870 nm, with a maximum irradiance of 1 mW/cm² at the plane of the eye. There are no lasers in the headgear. ETVision carries FCC Part 15 and CE marking, and the full list of standards, with editions, is in section 2.2 of the ETVision Manual.
Yes. Argus Science carries out optical, mechanical and algorithm modifications in house, and has integrated ETVision into helmets and special types of eyewear for specific projects. Tell us the headgear and the task, and the engineers who built the system will tell you what is possible.
A Windows 11 or Windows 10 Pro 64-bit PC with an Intel i7-8750H or better, 8 GB of memory and an NVIDIA RTX 2060 with 6 GB or better. For AI LMOT, 16 GB of memory and an RTX 3060 or better. The graphics card is the critical part; cards with 4 GB or less are not suitable. Argus Science can supply a laptop with the software installed.
For a long time. Support goes directly to the engineers who built the system, with training at installation and afterwards, and partners extend it worldwide. ETMobile and ETServer instruments are still serviced, and ETAnalysis opens recordings from ETVision and from earlier ETMobile, ETServer and ASL systems, so a laboratory keeps its history when it buys a new instrument.
Argus Science runs free demonstrations on a task close to your own, and a free trial of ETAnalysis with sample data is available on request. Outside the United States, you can buy directly from Argus Science or through one of the distributors listed on the Distributors page; either route works. Argus Science customers are on five continents, and Argus Science is ITAR compliant.
Both eyes measured independently at 180 Hz, worn over the participant's own glasses, with live object recognition included.
What the standard model recognizes, how to train your own objects, and what runs where.
Eight questions that decide which wearable eye tracker fits a study.
Tell us what you need to measure and in what environment. Demonstrations are free.
Schedule a demo