Pith. sign in

REVIEW 4 major objections 5 minor 86 references

VergeIO: Depth-Aware Eye Interaction on Glasses

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read A two-electrode EOG layout on glasses can sense vergence, the eyes' depth-focus movements, and classify up to six depth-based gestures with 77–98% accuracy across users without calibration, enabling hands-free depth-aware interaction.

desk verdict A real first for glasses-compatible EOG vergence sensing, but the headline numbers need careful reading and the vergence ground-truth gap should be addressed before acceptance. read the letter →

arxiv 2507.02187 v3 pith:744SEDYP submitted 2025-07-02 cs.HC

classification cs.HC
keywords vergenceelectrooculography(EOG)smartglassesdepth-awareinteractioneyegesturesdryelectrodesgesturerecognitionwearablesensing
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

VergeIO is an electrooculography (EOG)-based glasses system whose central claim is that depth, encoded in the eyes' vergence movements, can serve as an everyday hands-free input channel. The paper argues that a two-channel electrode layout on a normal glasses frame captures the opposing electrical potentials of vergence that earlier glasses layouts miss, and that this is enough to distinguish between four and six depth-based gestures. In a study across 20 users, personalized models reach 83–98% accuracy and a leave-one-user-out model reaches 77–97% with no calibration, all on dry electrodes at about 3 mW of sensing power. If correct, this makes depth of gaze, rather than only direction, a practical interaction primitive for smart glasses, varifocal lenses, and health monitoring.

What carries the argument

The load-bearing object is the two-channel temple-and-nose-bridge electrode geometry, which creates the large spatial separation needed to register the small, opposing vergence potentials instead of canceling them. Around it sit three supporting mechanisms: a motion-artifact pipeline, a logistic regression over statistical features of 2-second windows, that rejects facial and body-movement segments; a preamble-based activation scheme in which a brow-raise arms the system, cutting false positives to 0% across static, walking, chewing, and talking conditions; and a classification chain of Savitzky–Golay smoothing, 30 mV peak detection, and a random forest over ten features per channel. The angle-spacing choice, vergence angles of 9.5°, 4.1°, and 1.4° giving pairwise differences of 2.7°, 5.4°, and 8.1°, makes the three distances separable in EOG terms.

What would settle it

Record EOG from the same electrode layout while a head-fixed participant follows a bead that moves in depth, with a camera-based eye tracker as ground truth: if traces labeled as vergence appear without actual vergence, or if removing head-motion trials drops classification accuracy to chance, the depth-awareness claim is refuted.

Watch

Extended reading notes

Core claim

The paper's discovery is that vergence is detectable from a glasses-compatible EOG configuration when the electrodes are placed to capture the asymmetric, opposite-polarity signals the two eyes produce during convergence and divergence. By placing positive electrodes at the temples, a reference on the nose bridge, and ground at the mastoid, VergeIO records distinct waveforms for shifts among near (30 cm), mid (70 cm), and far (200 cm) viewing distances. A random-forest classifier on 20 features per gesture achieves 82.68% accuracy over six gestures within sessions and 77.36% across unseen users; a high-separability subset of four gestures reaches 97.99% within session and 97.43% across users, and a model trained on real-world Brock-string cues transfers to stereoscopic virtual cues and approximate thumb-based cues.

Load-bearing premise

The study labels each EOG trace as a vergence event based on instructions to focus on a Brock-string bead, without an independent eye-tracker confirming that the eyes actually converged or diverged rather than the head, lens, or facial muscles producing the signal.

Editorial extensions

If this is right

  • Varifocal eyeglasses can use vergence events to switch focal power among near, intermediate, and far zones without a camera or active depth sensing.
  • Augmented-reality glasses can treat a shift of gaze depth as selection of a virtual object rendered at a different stereoscopic depth, even when the display is physically at a fixed distance.
  • Safety monitoring can detect abrupt gaze-depth shifts, such as from the road to a phone, using only the EOG signal already available on the glasses frame.
  • The four-gesture subset, with cross-user accuracy above 97%, offers a deployable interaction vocabulary that works without per-user calibration or enrollment.
  • Remote screening for conditions such as convergence insufficiency can be driven by comparing EOG traces against approximate depth cues like thumbs and a distant object.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The four-gesture accuracy numbers suggest the deployable core of the system is the high-separability subset; the six-gesture confusions among adjacent mid-range transitions may reflect a real limit of EOG spatial resolution rather than an engineering fix.
  • A decisive test not yet run is to record the same electrode layout with a head-fixed participant and an independent eye-tracking ground truth, separating genuine vergence from accommodation-linked or head-motion confounds; the paper's cross-user claims would be on firmer ground if the classifier still separates targets under that protocol.
  • The paper's motion-artifact and preamble results hint that an implicit activation mechanism, such as fusing the EOG stream with a lightweight inertial sensor, could replace the brow-raise preamble and preserve the near-zero false-positive rate without interrupting interaction flow.
  • The uniform-vergence-angle design rule for choosing distances is generic: other triples of daily distances could be substituted for 30, 70, and 200 cm for specialized applications such as industrial work zones or cockpit displays, and the classifier should transfer as long as the pairwise angle differences remain comparable.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper presents VergeIO, an EOG-based smart-glasses prototype with dry electrodes at the temples and nose bridge, designed to detect depth-dependent vergence eye movements. The system defines six vergence gestures between 30, 70, and 200 cm, and also evaluates a reduced four-gesture subset chosen for separability. In a 20-participant study, the authors report 82.68% within-session accuracy for six gestures and 97.99% for four gestures with personalized models, with cross-session accuracy of 77.78% and cross-user accuracy of 77.36% (six gestures) or 97.43% (four gestures) without calibration. Additional contributions include a motion artifact detection pipeline, a brow-raise preamble to reduce false positives, a varifocal glasses application, and an open-source hardware/software plan.

Significance. If the underlying eye-movement mechanism is verified, the contribution is valuable: a glasses-compatible electrode configuration that captures the asymmetric EOG potentials of vergence would expand the interaction vocabulary of EOG glasses, and the hardware is simple, low-cost, dry-electrode, and low-latency. The paper is notably transparent about performance drops after remounting and across users, and it includes a direct comparison with the JINS MEME electrode layout. The 82.68% six-gesture within-session result is a genuine empirical contribution, and the 96.1% accuracy on traditional eye-movement classes shows the design does not sacrifice compatibility. However, the central claim that the system senses vergence rather than a correlated confound is not independently validated, and the headline 97% cross-user figure comes from a post hoc four-gesture subset evaluated on only 11 of the 20 participants. These issues make the significance of the main claim conditional.

major comments (4)
  1. [Sec. 4.2, Sec. 5.1, Sec. 7.5]
  2. [Sec. 5.1, Abstract, Conclusion]
  3. [Sec. 5.1, Table 3, Fig. 12]
  4. [Sec. 5.9, Abstract]
minor comments (5)
  1. [Sec. 5.4, Contributions]
  2. [Table 4]
  3. [Abstract vs. Sec. 4.2]
  4. [Sec. 3.4, Sec. 5.7]
  5. [Sec. 5.5, Sec. 5.6]

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the system is evaluated empirically with held-out sessions and users; no load-bearing derivation reduces to its own inputs.

full rationale

VergeIO is an empirical systems paper rather than an analytic derivation, so most circularity patterns do not apply. The central claim—that a two-channel temple/nose-bridge EOG layout can classify vergence gestures—is supported by a 20-participant user study with within-session, cross-session, and leave-one-user-out evaluations. Classifier labels come from the Brock String protocol (target bead fixated), not from the EOG signal itself, so the classification result is not definitionally equivalent to the input signal. The 30 mV peak-detection threshold is described as determined empirically from a pilot study, and the reported four-gesture subset is explicitly presented as a set with high separability; these are design and evaluation choices that affect optimism, but they are not fitted parameters renamed as predictions. The paper contains no load-bearing self-citations by the current authors, and its comparison against JINS MEME rests on an external prior study and a same-participant SNR/accuracy comparison. The lack of an independent eye-tracking ground truth for verifying that labeled events are genuine vergence is a substantive validity concern about confounds, not a circularity concern, because the paper does not define vergence in terms of the EOG outcome. The limitations section acknowledges the controlled-setting scope and cross-user/cross-session drops, which further supports that the reported numbers are empirical results rather than constructed equivalences.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The system does not introduce new physical entities; its central claim rests on physiological assumptions (EOG proportional to vergence angle, Brock String labels as ground truth), a single-participant pilot for the electrode design, and hand-chosen parameters including the 30 mV threshold, the MAD multiplier, the distance set, and the four-gesture subset.

free parameters (4)
  • Peak detection amplitude threshold = 30 mV
    Set empirically from the pilot study in Sec 3.5 to detect vergence onsets; it controls which signal peaks become classification segments, so it directly affects reported accuracy.
  • Vergence distance set = 30, 70, 200 cm (vergence angles 9.5, 4.1, 1.4 degrees)
    Chosen in Sec 3.1 to produce uniformly spaced vergence-angle differences (2.7, 5.4, 8.1 degrees) so signals are distinguishable; the task definition determines the classification boundaries.
  • Silence-removal MAD multiplier = 4.5
    Heuristic multiplier applied to the median absolute deviation of the first 2 seconds of each channel in Sec 3.4; it decides which windows are processed by the classifier.
  • Four-gesture subset selection = 30 to 200, 70 to 200, 200 to 30, 200 to 70 cm
    Post hoc selection of the four most separable gestures in Sec 5.1; the headline 97% accuracy refers to this subset, not the full six-gesture task.
assumptions (4)
  • domain assumption EOG signal amplitude is proportional to changes in vergence angle.
    Invoked in Sec 3.1 and Fig 3a to justify choosing distances by vergence angle and interpreting the recorded waveforms as vergence movement.
  • domain assumption The Brock String protocol induces genuine vergence movements in participants.
    The study labels EOG segments as vergence based on the clinical protocol in Sec 4.2 without independent eye-tracking ground truth; confounds such as head motion or accommodative muscle activity could corrupt labels.
  • domain assumption The single-participant SNR comparison generalizes to the 20-user study population.
    The 8.9 dB vs 1.6 dB SNR advantage over JINS MEME (Sec 3.2) was measured on one participant and motivates the electrode design and the expected separability in the full study.
  • domain assumption An interpupillary distance of 50 mm represents adult users for vergence-angle calculations.
    Used in Sec 3.1 to map distances to angles; adult IPD ranges from 50 to 75 mm, so true angular differences vary across participants.

how reviews work

0 comments
Cite this review

Pith. "Pith review of VergeIO: Depth-Aware Eye Interaction on Glasses." pith.science (2026). https://pith.science/paper/744SEDYP

@misc{pith2026250702187,
  author       = {Pith},
  title        = {Pith review of: VergeIO: Depth-Aware Eye Interaction on Glasses},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/744SEDYP}},
  note         = {Machine review of arXiv:2507.02187}
}
read the original abstract

There is growing industry interest in unobtrusive designs for electrooculography (EOG) sensing of eye gestures on glasses (e.g. JINS MEME and Apple eyewear). We present VergeIO, an EOG-based glasses system that enables depth-aware eye interaction by sensing vergence with a glasses-compatible electrode layout and smart glass prototype. It can distinguish between four depth-based eye gestures with 97% accuracy on unseen users without any calibration in a user study across 20 users and 1,520 gesture instances. To reduce false detections, we incorporate a motion artifact detection pipeline and a preamble-based activation scheme. The system uses dry sensors without any adhesives or gel and operates in real time with 3 mW power consumption by the analog sensing front-end.

Figures

Figures reproduced from arXiv: 2507.02187 by the authors.

Figure 1
Figure 1. Overview of the VergeIO system. (a) VergeIO senses eye vergence on glasses using electrooculography (EOG) by detecting gaze shifts between objects at different depths. The eyes diverge when shifting focus from a near to a far object, and converge when focusing from a far to a near object. (b) Our system can distinguish six vergence shifts between three everyday vergence distances. Each shift corresponds to a unique … view at source ↗
Figure 2
Figure 2. Applications of VergeIO. (a) Varifocal lenses adjust focal depth as the user shifts gaze between ambient reference points: a nearby phone and a distant TV. (b) Smart glasses enable gaze-based selection between virtual objects at different depths. (c) Remote screening of ocular disorders using approximate reference points: thumbs and a distant object. Given their proximity to the eyes, glasses are uniquely positioned… view at source ↗
Figure 3
Figure 3. Geometric characterization and signal validation for vergence distance selection. (a) VergeIO measures changes in vergence angle (𝜃) using EOG signals, which vary proportionally with these changes. (b) Our system can distinguish changes in vergence angles corresponding to three interaction distances: 30, 70, and 200 cm. (c) Mean EOG signal for each of the six vergence gestures for the left and right eye. Shaded regi… view at source ↗
Figures from the paper (17 more)
Figure 4
Figure 4. Figure 4: Limitations of prior EOG electrode placements on glasses. (a) Design of JINS MEME [11] and GAPses [32] lacks spatial separation which hinders vergence detection. Source: https://www.ohmyglasses.jp/blog/2016/05/31/jins-meme-new￾world/ (b) VergeIO places electrodes at th…
Figure 5
Figure 5. Figure 5: SNR comparison across six vergence gestures. VergeIO outperforms the commercial JINS MEME glasses [11]. To address these limitations, our system is designed to support vergence sensing while being compatible with a glasses form factor. Our design (Fig. 4b) places posit…
Figure 6
Figure 6. Figure 6: Hardware design. (a) Exploded CAD design illustrating the different components of the hardware design. (b) Fabricated design annotated with location of different components. (c) Front-view of glasses with dimensions. 3.3 Hardware design We create a glasses hardware pro…
Figure 7
Figure 7. Figure 7: VergeIO processing pipeline. Cost. The off-the-shelf cost is $140, with the optimized cost $73. The cost of the ELDRY electrode coating (4 cm2 , 0.5 mm thickness) is $20; Li-Po rechargeable battery is $15; Li-Po micro charger is $6; microcontroller development board is…
Figure 8
Figure 8. Figure 8: EOG signals recorded from the left and right eye during vergence, motion and noise events. (a) Six vergence gestures. (b) Vergence is distinct from eye movements like saccades and blinking. (c) Facial movements produce higher amplitude signals than vergence alone. (d) …
Figure 9
Figure 9. Figure 9: Experiment timeline. Participants first underwent a setup phase to ensure the glasses were properly mounted and received training on performing vergence movements along the Brock String. During the experimental phase, they completed 10 rounds, each consisting of the si…
Figure 10
Figure 10. Figure 10: Experimental setup for user study. The stars denote the location of colored beads tied to the Brock String and used as distance markers for the vergence gestures. 4.2 Data collection protocol We began by fitting each participant with our glasses prototype. The flexibl…
Figure 11
Figure 11. Figure 11: (a) Classification accuracy (𝑛 = 20) for within-session (training and testing on data collected without remounting the glasses) versus cross-session (training on one session and testing on the unseen session after re-mounting glasses). P19 and P20 withdrew before sess…
Figure 12
Figure 12. Figure 12: (a) Confusion matrix of within-session 6 gestures classification across all the participants and sessions. (b) Confusion matrix of cross-session 6 gestures classification. (c) Confusion matrix of within-session 4 gestures classification. Within Session Cross Session C…
Figure 13
Figure 13. Figure 13: (a) Effect of training duration on classification accuracy. (b) Motion artifact removal performance across vergence, motion, and noise EOG segments. (c) False positive rate under different movement scenarios: static, walking, chewing, and talking. A preamble consistin…
Figure 14
Figure 14. Figure 14: Subgroup analysis on within-session classification accuracy for [PITH_FULL_IMAGE:figures/full_fig_p015_14.png]
Figure 15
Figure 15. Figure 15: Confusion matrix classifying between different vergence and non-vergence movements across 11 classes. [PITH_FULL_IMAGE:figures/full_fig_p017_15.png]
Figure 16
Figure 16. Figure 16: Experimental setup to evaluate generalization to [PITH_FULL_IMAGE:figures/full_fig_p018_16.png]
Figure 17
Figure 17. Figure 17: Demographic summary of participants in the user study. Survey data was collected on 𝑛 = 20 participants of the data collection procedure. 5.9 Power consumption We measured the power consumption of our system using the Monsoon High Voltage Power Monitor [14] at 3.7 V, …
Figure 18
Figure 18. Figure 18: User experience survey (𝑛 = 20). Histograms summarize user perceptions across (a) ease of maintaining electrode contact, (b) ease of facial muscle movement while wearing the prototype, (c) concern regarding accidental gestures triggers, (d) interaction-induced fatigue…
Figure 19
Figure 19. Figure 19: (a) Prototype of mechanically adjustable varifocal glasses with an integrated servo motor for actuation. (b) Experi￾mental setup to determine the effective focal length at each varifocal lens setting. (c) Effective focal length of the varifocal glasses for different k…
Figure 20
Figure 20. Figure 20: Stereoscopic depth rendering is achieved by offsetting the left and right images to simulate the three depth (30, 70, [PITH_FULL_IMAGE:figures/full_fig_p022_20.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

86 extracted references · 71 canonical work pages

  1. [1]

    2PCS SG90 Servo Motor Micro Servo 9G Servo Motor for RC Robot Arm Helicopter Airplane Remote Control (2X)

    2026. 2PCS SG90 Servo Motor Micro Servo 9G Servo Motor for RC Robot Arm Helicopter Airplane Remote Control (2X). (2026). https://www.amazon.com/Servo-Helicopter-Airplane-Remote-Control/dp/B09L525KDT/ref=asc_df_B09L525KDT

  2. [2]

    Adafruit Micro-Lipo Charger for LiPoly Batt with USB Type C Jack

    2026. Adafruit Micro-Lipo Charger for LiPoly Batt with USB Type C Jack. (2026). https://www.adafruit.com/product/4410

  3. [3]

    Amazon Echo Frames

    2026. Amazon Echo Frames. (2026). https://www.amazon.com/b?ie=UTF8&node=117191383011

  4. [4]

    BioAmp EXG Pill

    2026. BioAmp EXG Pill. (2026). https://www.digikey.com/en/products/detail/upside-down-labs/0047/19525716

  5. [5]

    2026. Elcyo. (2026). https://en.elcyo.com/

  6. [6]

    Eye detection methods and devices

    2026. Eye detection methods and devices. (2026). https://patents.google.com/patent/US20240134449A1/en?oq=US20240134449

  7. [7]

    Flex Focal Adjustable Glasses Dial Vision Reading Glasses 0D to 350D Zoom Eyeglasses Men and Women

    2026. Flex Focal Adjustable Glasses Dial Vision Reading Glasses 0D to 350D Zoom Eyeglasses Men and Women. (2026). https: //www.amazon.com/PUDOEN-Adjustable-Glasses-Reading-Eyeglasses/dp/B0GD5PP3XV/

  8. [8]

    Google Glass review

    2026. Google Glass review. (2026). https://www.techradar.com/reviews/gadgets/google-glass-1152283/review/7

Show all 86 references
  1. [9]

    Google Glass smart eyewear returns

    2026. Google Glass smart eyewear returns. (2026). https://www.bbc.com/news/technology-40644195

  2. [10]

    IXI – The world’s first autofocus eyewear

    2026. IXI – The world’s first autofocus eyewear. (2026). https://ixieyewear.com/

  3. [11]

    JINS MEME

    2026. JINS MEME. (2026). https://jinsmeme.com/en/. , Vol. 1, No. 1, Article . Publication date: February 2026. 24•Xiyuxing Zhang, Duc Vu, Chengyi Shen, Zhikai Qin, Yuntao Wang, Yuanchun Shi, and Justin Chan

  4. [12]

    Learn about IPD and lens spacing on Meta Quest

    2026. Learn about IPD and lens spacing on Meta Quest. (2026). https://www.meta.com/help/quest/261777072346131/

  5. [13]

    Meta Quest 2

    2026. Meta Quest 2. (2026). https://www.meta.com/quest/products/quest-2/

  6. [14]

    Monsoon High Voltage Power Monitor

    2026. Monsoon High Voltage Power Monitor. (2026). https://www.msoon.com/high-voltage-power-monitor

  7. [15]

    Pupil Lab Neon Eye Tracking Glasses

    2026. Pupil Lab Neon Eye Tracking Glasses. (2026). https://pupil-labs.com/products/neon

  8. [16]

    Teensy 4.1

    2026. Teensy 4.1. (2026). https://www.sparkfun.com/teensy-4-1.html

  9. [17]

    Thin Lens Equation

    2026. Thin Lens Equation. (2026). https://openstax.org/books/university-physics-volume-3/pages/2-4-thin-lenses

  10. [18]

    Tobii Pro Glasses 3

    2026. Tobii Pro Glasses 3. (2026). https://www.tobii.com/products/eye-trackers/wearables/tobii-pro-glasses-3

  11. [19]

    ViXion01

    2026. ViXion01. (2026). https://vixion.jp/vixion01/?wovn=en

  12. [20]

    Aws M Abdullah, Ali R Ibrahim, Ammar A Al-Hamadani, Mohammed K Al-Obaidi, and Anas F Ahmed. 2023. A review for filtering techniques of the Electrooculography (EOG) signals.Glob. J. Eng. Technol. Adv16 (2023), 163–171

  13. [21]

    Deepshikha Acharya, Asha Rani, Shivangi Agarwal, and Vijander Singh. 2016. Application of adaptive Savitzky–Golay filter for EEG signal processing.Perspectives in Science8 (2016), 677–679. https://doi.org/10.1016/j.pisc.2016.06.056 Recent Trends in Engineering and Material Sciences

  14. [22]

    Sunggeun Ahn, Jeongmin Son, Sangyoon Lee, and Geehyuk Lee. 2020. Verge-it: Gaze interaction for a binocular head-worn display using modulated disparity vergence eye movement. InExtended abstracts of the 2020 CHI conference on human factors in computing systems. 1–7

  15. [23]

    2020.PyCaret: An open source, low-code machine learning library in Python

    Moez Ali. 2020.PyCaret: An open source, low-code machine learning library in Python. https://www.pycaret.org PyCaret version 1.0

  16. [24]

    Rafael Barea, Luciano Boquete, Manuel Mazo, and Elena López. 2002. System for assisted mobility using eye movements based on electrooculography.IEEE transactions on neural systems and rehabilitation engineering10, 4 (2002), 209–218

  17. [25]

    Chama Belkhiria, Atlal Boudir, Christophe Hurter, and Vsevolod Peysakhovich. 2022. Eog-based human–computer interface: 2000–2020 review.Sensors22, 13 (2022), 4914

  18. [26]

    Laurent Blondé, Didier Doyen, and Thierry Borel. 2010. 3D stereo rendering challenges and techniques. In2010 44th Annual Conference on Information Sciences and Systems (CISS). IEEE, 1–6

  19. [27]

    Yingxi Chen and Wyatt S Newman. 2004. A human-robot interface based on electrooculography. InIEEE International Conference on Robotics and Automation, 2004. Proceedings. ICRA’04. 2004, Vol. 1. IEEE, 243–248

  20. [28]

    Yun-Hsuan Chen, Maaike Op de Beeck, Luc Vanderheyden, Evelien Carrette, Vojkan Mihajlović, Bernard Grundlehner, Stefanie Gadeyne, Paul Boon, and Chris Van Hoof. 2014. Soft, Comfortable Polymer Dry Electrodes for high Quality ECG and EEG Recording.Sensors (Basel, Switzerland)14...

  21. [29]

    Chul Woo Cho, Ji Woo Lee, Kwang Yong Shin, Eui Chul Lee, Kang Ryoung Park, Heekyung Lee, and Jihun Cha. 2012. Gaze Detection by Wearable Eye-Tracking and NIR LED-Based Head-Tracking Device Based on SVR.Etri Journal34, 4 (2012), 542–552

  22. [30]

    Neil A Dodgson. 2004. Variation and extrema of human interpupillary distance. InStereoscopic displays and virtual reality systems XI, Vol. 5291. SPIE, 36–46

  23. [31]

    2021.detecta: A Python module to detect events in data

    Marcos Duarte. 2021.detecta: A Python module to detect events in data. https://doi.org/10.5281/zenodo.4598962

  24. [32]

    Sebastian Frey, Mattia Alberto Lucchini, Victor Kartsch, Thorir Mar Ingolfsson, Andrea Helga Bernardi, Michael Segessenmann, Jakub Osieleniec, Simone Benatti, Luca Benini, and Andrea Cossettini. 2024. GAPses: Versatile smart glasses for comfortable and fully-dry acquisition an...

  25. [33]

    Michal Gnacek, Ifigeneia Mavridou, John Broulidakis, Mohsen Fatoorechi, Ellen Seiss, Theodoros Kostoulas, Emili Balaguer-Ballester, Ivana Kiprijanovska, Claire Rosten, and Charles Nduka. 2022. emteqPRO—Fully Integrated Biometric Sensing Array for Non-Invasive Biomedical Resear...

  26. [34]

    Eugene Hamori, Roy D Broad, and Calvin E Reed Jr. 1982. Study of unaided cross-eyed stereopsis.Perception11, 3 (1982), 297–304

  27. [35]

    Layali Ibrahim Hassan, Samira Mohamed Ibrahim, and Mustafa Abdu. 2017. Efficacy of home-based vision therapy for convergence insufficiency in secondary schools’ students.Sudanese Journal of Ophthalmology9, 1 (2017), 5–9

  28. [36]

    Patrick Heijden, Camille Gilbert, Samira Jafari, and Mattia Lucchini. 2024. Multi-Channel Soft Dry Electrodes for Electrocardiography Acquisition in the Ear Region.Sensors24 (01 2024), 420. https://doi.org/10.3390/s24020420

  29. [37]

    Brien A Holden, Timothy R Fricke, S May Ho, Reg Wong, Gerhard Schlenther, Sonja Cronjé, Anthea Burnett, Eric Papas, Kovin S Naidoo, and Kevin D Frick. 2008. Global vision impairment due to uncorrected presbyopia.Archives of ophthalmology126, 12 (2008), 1731–1739

  30. [38]

    Mark S. Hughes. 1991. Dictionary of Eye Terminology.Archives of Ophthalmology109, 9 (09 1991), 1208–1208. https://doi.org/10.1001/ archopht.1991.01080090032015 arXiv:https://jamanetwork.com/journals/jamaophthalmology/articlepdf/639252/archopht_109_9_015.pdf

  31. [39]

    Shoya Ishimaru, Kai Kunze, Yuji Uema, Koichi Kise, Masahiko Inami, and Katsuma Tanaka. 2014. Smarter eyewear: Using commercial EOG glasses for activity recognition. InProceedings of the 2014 ACM International Joint Conference on Pervasive and Ubiquitous Computing: Adjunct Publ...

  32. [40]

    Robert J. K. Jacob. 1990. What you look at is what you get: eye movement-based interaction techniques. InProceedings of the SIGCHI Conference on Human Factors in Computing Systems(Seattle, Washington, USA)(CHI ’90). Association for Computing Machinery, New York, NY, USA, 11–18...

  33. [41]

    Elizabeth Carolina Jiménez, August Romeo, Laura Pérez Zapata, Maria Solé Puig, Patricia Bustos-Valenzuela, José Cañete, Paloma Varela Casal, and Hans Supèr. 2020. Eye vergence responses in children with and without reading difficulties during a word detection task. Vision Rese...

  34. [42]

    Kanoh, S

    S. Kanoh, S. Ichi-nohe, S. Shioya, K. Inoue, and R. Kawashima. 2015. Development of an eyewear to measure eye and body movements. In2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). 2267–2270. https: //doi.org/10.1109/EMB...

  35. [43]

    Minsol Kim, Serena Pei, Samiksha Singh, Selam Brook, and Nataliya Kosmyna. 2024. Privacy-Preserving Eye Movement Classification with Camera-Free, Non-Invasive EOG-EEG Glasses. In2024 IEEE 20th International Conference on Body Sensor Networks (BSN). IEEE, 1–4

  36. [44]

    Tobias King, Michael Knierim, Philipp Lepold, Christopher Clarke, Hans Gellersen, Michael Beigl, and Tobias Röddiger. 2025. earEOG via Periauricular Electrodes to Facilitate Eye Tracking in a Natural Headphone Form Factor.arXiv preprint arXiv:2506.07193(2025)

  37. [45]

    Dominik Kirst and Andreas Bulling. 2016. On the verge: Voluntary convergences for accurate and precise timing of gaze input. In Proceedings of the 2016 CHI Conference Extended Abstracts on Human Factors in Computing Systems. 1519–1525

  38. [46]

    Nataliya Kosmyna, Karim Adl, and Minsol Kim. 2024. Wearable Pair of EEG, EOG and fNIRS Glasses for Cognitive Workload Detection. 1–4. https://doi.org/10.1109/BSN63547.2024.10780518

  39. [47]

    Nataliya Kosmyna, Caitlin Morris, Utkarsh Sarawgi, Thanh Nguyen, and Pattie Maes. 2019. AttentivU: A wearable pair of EEG and EOG glasses for real-time physiological processing. In2019 IEEE 16th International Conference on Wearable and Implantable Body Sensor Networks (BSN). IEEE, 1–4

  40. [48]

    Nataliya Kosmyna, Utkarsh Sarawgi, and Pattie Maes. 2018. AttentivU: Evaluating the feasibility of biofeedback glasses to monitor and improve attention. InProceedings of the 2018 ACM International Joint Conference and 2018 International Symposium on Pervasive and Ubiquitous Co...

  41. [49]

    Shinya Kudo, Hiroyuki Okabe, Taku Hachisu, Michi Sato, Shogo Fukushima, and Hiroyuki Kajimoto. 2013. Input method using divergence eye movement. InCHI’13 Extended Abstracts on Human Factors in Computing Systems. 1335–1340

  42. [50]

    Joong Hoon Lee, HanSeop Kim, Ji-Young Hwang, Jinmook Chung, Tae-Min Jang, Dong Seo, Yuyan Gao, Junhyun Lee, Haedong Park, Seungwoo Lee, Hong Moon, Huanyu Cheng, Sang-Hoon Lee, and Suk-Won Hwang. 2020. 3D Printed, Customizable and Multi-functional Smart Electronic Eyeglasses (E...

  43. [51]

    Philipp Lepold, Tobias Röddiger, Tobias King, Kai Kunze, Christoph Maurer, and Michael Beigl. 2024. OpenEarable ExG: Open-Source Hardware for Ear-Based Biopotential Sensing Applications. InCompanion of the 2024 on ACM International Joint Conference on Pervasive and Ubiquitous ...

  44. [52]

    Ke Li, Ruidong Zhang, Boao Chen, Siyuan Chen, Sicheng Yin, Saif Mahmud, Qikang Liang, François Guimbretière, and Cheng Zhang

  45. [53]

    Tianxing Li and Xia Zhou. 2018. Battery-free eye tracker on glasses. InProceedings of the 24th Annual International Conference on Mobile Computing and Networking. 67–82

  46. [54]

    Peichen Liu, Sadasivan Puthusserypady, I Scott MacKenzie, Cihan Uyanik, and John Paulin Hansen. 2025. EarEOG: Using Headphones and Around-the-Ear EOG Signals for Real-Time Wheelchair Control.Proceedings of the ACM on Human-Computer Interaction9, 3 (2025), 1–16

  47. [55]

    Alberto Lopez, Francisco J Ferrero, Marta Valledor, Juan C Campo, and Octavian Postolache. 2016. A study on electrode placement in EOG systems for medical applications. In2016 IEEE International symposium on medical measurements and applications (MeMeA). IEEE, 1–5

  48. [56]

    Ruichun Ma, Yasuo Morimoto, John S Ho, Sam Shiu, and Jiang Zhu. 2025. mmET: mmWave Radar-Based Eye Tracking on Smart Glasses. InProceedings of the 23rd ACM Conference on Embedded Networked Sensor Systems. 30–42

  49. [57]

    Gary L Mancil, Ian L Bailey, Kenneth E Brookman, J Bart Campbel, Michael H Cho, Alfred A Rosenbloom, and James E Sheedy. 2011. Optometric clinical practice guideline: Care of the patient with presbyopia.St. Louis, MO: American Optometric Association(2011)

  50. [58]

    Addison Mayberry, Pan Hu, Benjamin Marlin, Christopher Salthouse, and Deepak Ganesan. 2014. iShadow: design of a wearable, real-time mobile gaze tracker. InProceedings of the 12th annual international conference on Mobile systems, applications, and services. 82–94

  51. [59]

    Addison Mayberry, Yamin Tun, Pan Hu, Duncan Smith-Freedman, Deepak Ganesan, Benjamin M Marlin, and Christopher Salthouse

  52. [60]

    LN McLin and CM Schor. 1988. Voluntary effort as a stimulus to accommodation and vergence.Investigative ophthalmology & visual science29, 11 (1988), 1739–1746

  53. [61]

    Saswat Mishra, Yun-Soung Kim, Jittrapol Intarasirisawat, Young-Tae Kwon, Yongkuk Lee, Musa Mahmood, Hyo-Ryoung Lim, Robert Herbert, Ki Jun Yu, Chee Siang Ang, et al. 2020. Soft, wireless periocular wearable electronics for real-time detection of eye vergence in a virtual reali...

  54. [62]

    F Mizuno, T Hayasaka, K Tsubota, S Wada, and T Yamaguchi. 2003. Development of hands-free operation interface for wearable computer-hyper hospital at home. InProceedings of the 25th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (IEEE C...

  55. [63]

    Barath S Narayan, Sreyas Janamanchi, Achintya Harsha, Prashanth Jonna, Madhav Rao, et al. 2024. EOG-Assist: Real-Time EOG-Based Eye Tracking for Assistive Systems. In2024 18th International Conference on Control, Automation, Robotics and Vision (ICARCV). IEEE, 795–800

  56. [64]

    Nitish Padmanaban, Robert Konrad, and Gordon Wetzstein. 2018. Autofocals: gaze-contingent eyeglasses for presbyopes. InACM SIGGRAPH 2018 Emerging Technologies. 1–2

  57. [65]

    Bait, Juyoung Lee, Jingjing Xu, Roshan L Peiris, Woontack Woo, Mark Billinghurst, and Kai Kunze

    Yun Suen Pai, Marsel L. Bait, Juyoung Lee, Jingjing Xu, Roshan L Peiris, Woontack Woo, Mark Billinghurst, and Kai Kunze. 2022. NapWell: An EOG-based Sleep Assistant Exploring the Effects of Virtual Reality on Sleep Onset.Virtual Real.26, 2 (June 2022), 437–451. https://doi.org...

  58. [66]

    David W Patmore and R Benjamin Knapp. 1998. Towards an EOG-based eye tracker for computer control. InProceedings of the third international ACM conference on Assistive technologies. 197–203

  59. [67]

    DF Roberts, KA Provins, and RJ Morton. 1959. Arm strength and body dimensions.Human Biology31, 4 (1959), 334–343

  60. [68]

    Soha Rostaminia, Alexander Lamson, Subhransu Maji, Tauhidur Rahman, and Deepak Ganesan. 2019. W!nce: Unobtrusive sensing of upper facial action units with eog-based eyewear.Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies3, 1 (2019), 1–26

  61. [69]

    Lingyan Ruan, Bin Chen, and Miu-Ling Lam. 2018. Human-computer interaction by voluntary vergence control. InSIGGRAPH Asia 2018 Posters. 1–2

  62. [70]

    John T. Rule. 1941. The Geometry of Stereoscopic Projection.J. Opt. Soc. Am.31, 4 (Apr 1941), 325–334. https://doi.org/10.1364/JOSA.31. 000325

  63. [71]

    Wayne J Ryan, Andrew T Duchowski, and Stan T Birchfield. 2008. Limbus/pupil switching for wearable eye tracking under variable lighting conditions. InProceedings of the 2008 symposium on Eye tracking research & applications. 61–64

  64. [72]

    Abraham Savitzky and Marcel JE Golay. 1964. Smoothing and differentiation of data by simplified least squares procedures.Analytical chemistry36, 8 (1964), 1627–1639

  65. [73]

    Nicolas Schärer, Federico Villani, Aishwarya Melatur, Steven Peter, Tommaso Polonelli, and Michele Magno. 2024. ElectraSight: Smart Glasses with Fully Onboard Non-Invasive Eye Tracking Using Hybrid Contact and Contactless EOG.arXiv preprint arXiv:2412.14848 (2024)

  66. [74]

    Nicolas Schärer, Federico Villani, Aishwarya Melatur, Steven Peter, Tommaso Polonelli, and Michele Magno. 2025. Electrasight: Fully onboard eye tracking for smart glasses with hybrid eog (heog).IEEE Internet of Things Journal(2025)

  67. [75]

    Scheiman and B

    M. Scheiman and B. Wick. 2008.Clinical Management of Binocular Vision: Heterophoric, Accommodative, and Eye Movement Disorders. Wolters Kluwer Health/Lippincott Williams & Wilkins. https://books.google.com/books?id=jGGROHBFYt8C

  68. [76]

    Vargo, Aman Gupta, George Chernyshov, Kai Kunze, and Tilman Dingler

    Benjamin Tag, Andrew W. Vargo, Aman Gupta, George Chernyshov, Kai Kunze, and Tilman Dingler. 2019. Continuous Alertness Assessments: Using EOG Glasses to Unobtrusively Monitor Fatigue Levels In-The-Wild. InProceedings of the 2019 CHI Conference on Human Factors in Computing Sy...

  69. [77]

    Domenico Tringali, Dorian Haci, Federico Mazza, Konstantin Nikolic, Danilo Demarchi, and Timothy G Constandinou. 2021. Eye Accommodation Sensing for Adaptive Focus Adjustment. In2021 43rd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EM...

  70. [78]

    Christopher W Tyler and Maureen B Clarke. 1990. Autostereogram. InStereoscopic displays and applications, Vol. 1256. SPIE, 182–197

  71. [79]

    Christopher W Tyler, Anas M Elsaid, Lora T Likova, Navdeep Gill, and Spero C Nicholas. 2012. Analysis of human vergence dynamics. Journal of Vision12, 11 (2012), 21–21

  72. [80]

    Yuji Uema and Kazutaka Inoue. 2017. JINS MEME algorithm for estimation and tracking of concentration of users. InProceedings of the 2017 ACM International Joint Conference on Pervasive and Ubiquitous Computing and Proceedings of the 2017 ACM International Symposium on Wearable...

  73. [81]

    Ming Wan, Nan Zhang, Huamin Qu, and Arie E Kaufman. 2004. Interactive stereoscopic rendering of volumetric environments.IEEE Transactions on Visualization and Computer Graphics10, 1 (2004), 15–28

  74. [82]

    Hui-Shyong Yeo, Juyoung Lee, Woontack Woo, Hideki Koike, Aaron J Quigley, and Kai Kunze. 2021. JINSense: Repurposing Electroocu- lography Sensors on Smart Glass for Midair Gesture and Context Sensing. InExtended Abstracts of the 2021 CHI Conference on Human Factors in Computin...

  75. [83]

    Young and David Sheena

    Laurence R. Young and David Sheena. 1975. Survey of eye movement recording methods.Behavior Research Methods & Instrumentation 7, 5 (01 Sep 1975), 397–429. https://doi.org/10.3758/BF03201553

  76. [84]

    Chenyang Zhang, Tiansu Chen, Eric Shaffer, and Elahe Soltanaghai. 2024. FocusFlow: 3D Gaze-Depth Interaction in Virtual Reality Leveraging Active Visual Depth Manipulation. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems. 1–18. , Vol. 1, No. 1, A...

  77. [2015]

    InProceedings of the 21st Annual International Conference on Mobile Computing and Networking

    CIDER: Enabling robustness-power tradeoffs on a computational eyeglass. InProceedings of the 21st Annual International Conference on Mobile Computing and Networking. 400–412

  78. [2024]

    InProceedings of the 30th Annual International Conference on Mobile Computing and Networking

    GazeTrak: Exploring Acoustic-based Eye Tracking on a Glass Frame. InProceedings of the 30th Annual International Conference on Mobile Computing and Networking. 497–512

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.