In the past two years, the most “magical” change in the AR/XR industry has not been resolution or chips, but rather “how you talk to the glasses.” Apple’s Vision Pro made a stunning entrance with “eye tracking + pinch”; Meta bundled its Neural Band with Ray-Ban Display glasses; Meizu has already tested eye-tracking solutions on its StarV Explorer concept device. This article focuses on three core interaction methods:
- Eye Tracking Interaction
- EMG Wristband (Neural Band) Interaction
- Ring Interaction
We will explore everything from principles, pros and cons, technical maturity, cost & market price to which one is more likely to be the “ultimate interaction” in the long run, finally offering judgments on the future form of AR interaction. We apologize for any omissions in company statistics.
01 Principles and Experience Differences of the Three Interaction Methods
1. Eye Tracking Interaction: Point with “Looking,” Confirm with “Pinching”
Basic Idea: Infrared cameras + IR LEDs illuminate the eye area to capture pupil position and gaze vectors; the system calculates in real-time “where you are looking” and combines this with gestures/buttons to complete clicks, drags, etc.
Typical Products & Companies:
- Apple Vision Pro / Vision Pro 2: Heavily relies on “gaze + pinch” as the primary interaction; the latest M5 version still starts at $3,499.
- Magic Leap 2: Provides a complete eye-tracking API, using gaze for UI highlighting, selection, gaze rendering, etc.
- Pico Neo / Pico 4 Pro Eye: Integrates Tobii eye-tracking modules, supporting foveated rendering and behavior analysis.
- Samsung Galaxy XR and other next-gen MR headsets: All include eye tracking as a standard feature.
Pros:
- Extremely high pointing efficiency: Look where you want to click; naturally suited for “selecting targets.”
- High hardware integration: Cameras and IR modules are hidden within the optical system, requiring no extra wearables.
- Particularly useful for foveated rendering and user behavior analysis, saving power and enabling smarter UIs.
Cons:
- Cannot complete all interactions independently; must be paired with gestures/clicks (otherwise, “look and click” leads to severe accidental touches).
- Sensitive to wearing position, interpupillary distance, and glasses fit; requires calibration. Accuracy decreases for some groups (deep-set eyes, single eyelids, etc.).
- Privacy controversies: Gaze data + content preferences are golden data for behavioral profiling.
- If applied to AI/AR glasses, energy consumption is high; especially software-based eye-tracking solutions, power draw might be unacceptable for AI AR glasses.
Technical Maturity (Subjective Score): Approx. 8/10
Consumer products (Vision Pro, Magic Leap 2, Pico Pro series) are already commercialized with comprehensive developer documentation.
Cost & Price:
- Looking at BOM alone, the eye-tracking module (IR camera + LED + control board) cost is relatively low compared to the whole unit, but it increases optical/engine structural complexity.
- For the user, eye tracking is part of the entire headset/glasses:
- Vision Pro 2: Approx. $3,499 (includes eye tracking, gestures, whole system).
- Galaxy XR: Approx. $1,799, also integrating eye tracking.
2. EMG Wristband Interaction: From “Moving Muscles” to “Sending Commands”
Basic Principle: Arrange sEMG/SNC sensors on the wrist to collect weak electrical signals from forearm muscles or nerves, decoding them via models into actions like “thumb swipe, pinch, click, scroll,” or even “virtual typing.”
Representative Companies & Products:
- Meta:
- Meta Neural Band sold with the Ray-Ban Display smart glasses kit (kit approx. $799); lost bands can be purchased separately for approx. $199.
- Wearable Devices (Israel):
- Mudra Band / Mudra Link, targeting Apple Watch/XR devices. Mudra Link official price is $299, pre-order discount $199, focusing on touchless gesture control for XR/AR.
- Historical Product Thalmic Labs / Myo Armband: Discontinued, but iconic in the HCI field.
Pros:
- No reliance on cameras, no fear of occlusion: Works with hands in pockets or behind back; very suitable for outdoors and crowded places.
- Theoretically has a high bandwidth input ceiling: Can recognize independent movements of each finger, even realizing a wristband version of an “air keyboard.”
- Can be made into a watch strap form, coexisting with Apple Watch or regular watches—more natural experience.
Cons:
- Sensitive to wearing position, skin contact, sweat, etc.; requires calibration and individual training.
- Complex algorithm stack; high barrier to entry for small companies; slower ecosystem construction.
- Cost is higher than ordinary IMU wristbands; difficult to achieve “freebie-level” ultra-low prices in the short term.
Technical Maturity: Approx. 7–8/10
Major companies have mass-produced physical products (Ray-Ban Display + Neural Band), and C-end/XR products like Mudra Link have started shipping.
Cost & Market Price:
- Wristbands targeting XR/neural interaction mostly retail in the $200–300 range:
- Meta Neural Band: Separate replacement piece for $199.
- Mudra Link: Pre-order $199, original price $299.
3. Ring Interaction: Turning Fingers into “Remote Controls”
Basic Principle: Most ring interaction schemes use IMU (accelerometer + gyroscope) + touch/buttons + vibration, sending commands via finger posture and clicks. Some pair with arm trackers for more precise 6DoF.
Representative Companies & Solutions:
- Finch Technologies:
- FinchRing gesture ring, paired with upper arm tracker FinchTracker, providing 3DoF/6DoF gesture control for Nreal and other AR glasses.
- Huawei:
- Recent patent: Flexible temple straps for AR/VR glasses can be detached to become smart rings for pointing, navigation, and game interaction, following an “integrated design” approach.
- Early Universal Gesture Rings:
- Fin / Neyya, etc., using IMU + touchpads to control phones/PCs, historical prices around $100–150.
- Health Smart Rings (Oura, Samsung Galaxy Ring, etc.):
- Primarily for health monitoring and notifications, but prove the maturity of “ring-form hardware.”
Pros:
- Most discreet form factor, most like jewelry: Low psychological threshold for user acceptance; very suitable for daily wear.
- Suitable as an “enhanced remote control”: Click, swipe, scroll, point; executes simple and quick operations.
- Relatively simple BOM structure (IMU + MCU + BLE + small battery); hardware cost has room for compression.
Cons:
- Limited input bandwidth; difficult to achieve “text input” or “high-precision continuous control”; most schemes stay at the “remote control level.”
- Difficult finger size adaptation (complex sizing), easy to lose, inconvenient charging; small volume limits battery and antenna space.
- If relying on external visual tracking/trackers, still sensitive to environment and posture.
Technical Maturity: Approx. 5–6/10 (for AR interaction)
Health rings are very mature, but rings specifically for AR/VR gestures are still mainly B-end/DevKit/patent-focused, with few consumer-scale products.
Cost & Price:
- AR gesture rings (FinchRing) currently target B-end/developers; no large-scale public retail price yet.
- Referring to historical products like Neyya, generally $100–150.
- Health rings (Oura/Galaxy Ring, etc.) are mostly in the $250–400 range, serving as a price reference under similar volume and craftsmanship conditions.
02 Horizontal Comparison of Core Dimensions
2. Overview Table of Typical Companies & Products Worldwide
(Note: It is impossible to list all manufacturers; the table below focuses on companies closely related to AR/XR interaction or representative in the supply chain.)
03 Which is Closer to the “Ultimate Interaction Scheme for Future Smart Glasses”?
The short-term answer is: No single method can become the true “ultimate,” but if looking only at the dimension of “hand/upper limb wearable devices,” EMG wristbands have the highest ceiling; while eye tracking is almost a mandatory “first layer” for all high-end AR. Rings are more like a supplementary option.
1. From the Perspective of “Bandwidth”: EMG > Eye Tracking > Ring
- Eye Tracking: Very suitable for “selecting targets” and “scrolling pages,” but not suitable for continuous value adjustment or complex editing (still needs gestures/controllers).
- EMG Wristband: Has the potential to cover everything from “click, scroll” to “virtual keyboard input,” holding huge significance for heavy office work, content creation, gaming, etc.
- Ring: Better at “confirming,” “switching,” “simple direction control”; difficult to shoulder the role of “main input device.”
2. From the Perspective of “Binding with AR Routes”: Eye Tracking + EMG is the Main Line
- High-end headsets almost default to integrating eye tracking and continue to refine interaction experiences around “gaze + gestures” (Vision Pro, Magic Leap 2, Galaxy XR, etc.).
- Meta’s AI glasses + Neural Band combination is turning “neural wristband + glasses” into a complete product paradigm: users complete scrolling, clicking, rotating, etc., via the wristband.
- This shows: Eye tracking + some form of hand input (gestures/EMG/ring/controller) has become an industry consensus.
3. From the Perspective of “User Acceptance & Cost”: Eye Tracking is “Invisible Infrastructure,” EMG and Rings Have Their Own Positions
- User acceptance of “wearing an extra wristband” has actually been educated in the smartwatch era; as long as weight and wearing feel are OK, EMG wristbands have a chance to become regular wearables.
- The ring form has advantages in appearance, more suitable as:
- A light interaction entry point (silent clicks, shortcut operations)
- Or an integrated device for “health + notifications + light interaction.”
- Eye tracking is “formless yet superior”: As long as you buy high-end AR, it naturally exists. Users won’t pay separately for eye tracking, but will despise a product as “outdated” if it lacks it.
04 Predictions and Thoughts on AR Glasses Interaction Development in the Next 5–10 Years
Combining current product routes, I am more optimistic about the following evolutionary path:
1. Short Term (1–3 Years):
“Eye tracking + bare-hand gestures + voice” becomes standard for high-end; EMG/Rings land in specific scenarios first.
- High-end XR Headsets (Vision Pro series, Galaxy XR, etc.): Continue to strengthen eye-tracking accuracy, adding advanced interactions like “eye-controlled scrolling” and “gaze + voice.”
- AI Glasses: Lightweight AI glasses like Meta will use neural wristbands for differentiation, first capturing scenarios like “news reading + commuting + light office work.”
- Rings: Test the waters as accessories in scenarios like gaming, in-car, and meetings where “only silent clicks/swipes” are needed, and merge with health rings in some brands.
2. Mid Term (3–5 Years):
EMG wristbands enter more brand ecosystems; rings gradually evolve towards “health + light interaction.”
- More phone/glasses manufacturers will consider:
- Self-developing or cooperating to introduce neural wristbands (similar to Mudra Link’s “cross-platform controller”).
- Rings: Become one of the mainstream forms of “health bracelet + light input,” with AR interaction capabilities existing as an “additional selling point.”
3. Long Term (5–10 Years):
The true “ultimate interaction” is a multi-modal combination, not a single device.
A mature pair of AR glasses in the future will look more like this system:
- Eye Tracking: Responsible for “seeing who, selecting who,” achieving attention alignment and interface highlighting.
- EMG Wristband/Strap: Provides high-bandwidth, low-noise confirmation and continuous control, supporting fast text input, scrolling, graphic editing, and fine game control.
- Bare-hand Visual Gestures: Complete natural actions like grabbing, placing, and dragging within the field of view.
- Voice + Large Models: Express complex intentions in natural language (“summarize for me,” “change the PPT structure for me”), reducing operational burden.
- Rings / Controllers / Other Forms: Serve as supplements or replacements for specific scenarios (gaming, industrial, in-car, privacy scenarios).
If one must choose the “most likely to become the core hardware for future AR interaction” among the three:
- Eye tracking is the “foundation of all high-end AR”; almost no choice needed, it must be there.
- Among wearable input devices, EMG wristbands have the highest bandwidth and potential, with a chance to become the “main input layer” in professional and heavy-use scenarios.
- Rings are more like the lightest “auxiliary hand” in form, extremely valuable in daily/light scenarios, but difficult to rule alone.
05 Conclusion
Looking at these three AR interaction schemes—eye tracking, EMG wristbands, and smart rings—it feels like witnessing three different development threads:
- Eye tracking is the intuition of “what you see is what you get.”
- EMG is the ultimate secrecy of “mind moves, body stays still.”
- Rings are the pragmatism and convenience of “point and shoot.”
However, the road to the “ultimate interaction” has never been unique or certain. Should we pursue extreme naturalness and invisibility in interaction, or choose practicality for the present? Should machines understand our intentions, or should we adapt to the rule limitations of interaction tools? Perhaps there is no single winner in this race; the final answer is more likely a fusion of multi-modalities, with different technologies playing the leading role in different scenarios.
As a user, which interaction method excites you the most when using AR glasses in daily life? Is it the freedom of “eye-driven” control, the technological magic of the “mind-control wristband,” or the elegance and precision of “ring manipulation”?
In your opinion, will they eventually converge, or will one of them become dominant?
Welcome to leave your insights in the comments section!
About the Author
