WiFi Wireless Keyboard: How Neuralase Tracks Finger Movement Through Wireless Signals

Neuralase's WiFi Wireless Keyboard is a concept that reads standard WiFi signals to track finger position in real time, so you can type on an imaginary keyboard drawn on a sheet of paper. No physical keys, no wearable sensor. Each finger movement is meant to simulate a keystroke, turning ordinary WiFi infrastructure into a way of capturing typing motion.

Key Takeaways

• Neuralase's WiFi Wireless Keyboard concept reads WiFi signals to track finger position in real time, letting you type on an imaginary keyboard drawn on paper.
• Macquarie University's FingerDraw research reports a median finger-tracking accuracy of 1.27 cm using one WiFi transmitter and two receivers.
• Macquarie University's FingerDraw system recognized ten digits drawn in the air with over 93.0% average accuracy.
• EDN reported that the WiKey keystroke-recognition system reached accuracy as high as 97.5% under best-controlled conditions, dropping to 77.5% once other people were moving nearby.

How does WiFi track finger movements without a physical keyboard?

WiFi signals are electromagnetic waves, the same family of radiation as visible light, and they reflect and scatter off any surface they hit, including skin. As your fingers move, those reflections change in tiny, measurable ways. A receiver picks up those changes and a model translates them into motion.

The specific measurement most researchers rely on is called Channel State Information, or CSI. CSI describes how a WiFi signal's phase and amplitude shift as it travels from a transmitter to a receiver and bounces off objects in the room along the way. A hand held still produces one CSI pattern. A finger curling to tap an imaginary key produces a different, repeatable pattern. Once you have enough labelled examples of those patterns, you can train a model to recognize which key someone meant to press.

This is the same underlying physics we describe on wifi which walls where, where WiFi's ability to reflect, scatter, and diffract off materials like drywall, wood, and glass is what lets it sense a room instead of just connecting devices to it.

The research behind finger-tracking accuracy

Nobody at Neuralase is claiming to have invented finger tracking over WiFi from scratch. The idea has an academic track record, and it's worth being specific about what that record actually shows.

Macquarie University's FingerDraw paper, published in March 2020, used a single WiFi transmitter and two WiFi receivers to reconstruct finger-drawing trajectories, including digits, letters, and symbols. Macquarie University reports a median tracking accuracy of 1.27 cm across the system, and separately reports that recognizing ten digits traced in the air achieved an average accuracy of over 93.0%.

A different research effort, WiKey, focused specifically on keystroke recognition rather than freehand drawing. EDN reported in August 2016 that WiKey detected keystrokes with 93.5% accuracy in a controlled environment, using a Lenovo X200 laptop and a TP-Link WR1043ND router running the 802.11n/ac WiFi protocol. Under the best-controlled test conditions, EDN reported that WiKey's accuracy reached 97.5%. In a simulated real-world scenario with multiple people moving nearby, that figure dropped to 77.5%. That gap between lab conditions and a busy room is the honest, central challenge of this entire field, and it's one reason a working WiFi keyboard concept has to account for interference from other bodies in the same space, not just the typist's own hand.

What would a WiFi wireless keyboard concept actually look like to use?

You draw a keyboard layout on a piece of paper, or simply picture one on a flat surface, and type by moving your fingers as if the keys were really there. The Neuralase WiFi Wireless Keyboard model reads the surrounding WiFi signals and tracks where your fingers are, translating each tap into a keystroke.

The Neuralase WiFi Wireless Keyboard model reads your WiFi signals and can track exactly where your fingers are in real-time. Each finger movement can accurately simulate a key stroke, allowing you to use an imaginary keyboard.

The practical appeal is obvious for anyone who travels with a laptop and finds a full-size keyboard too bulky to carry. Instead of packing a separate keyboard, or squinting at a laptop's cramped keys on a train tray table, you'd rely on the WiFi already present in the room. You can read the fuller concept on the WIFI wireless keyboard page.

Comparing the research approaches

The table makes the trade-off visible. FingerDraw's centimetre-level precision suits drawn shapes and digits. WiKey's keystroke-specific training gets very high accuracy in a quiet room but loses almost twenty points of accuracy once other people are moving through the same space. Any WiFi keyboard concept has to be built with that real-world drop in mind, not just the best-case lab number.

Where WiFi sensing goes beyond keyboards

At Neuralase, we are exploring how existing wireless infrastructure could become more than a communications network. A wireless keyboard is one application of that idea, but the same CSI-based sensing sits behind several other projects we've published.

WIFI room volumetric capture uses 2.4GHz CSI across 16 channels to map the phase and amplitude of reflected waves and reconstruct a room's 3D structure. WIFI human dense pose applies related signal processing to estimate body pose from reflections alone. Both projects work from the same starting point as the keyboard concept: a WiFi signal that reflects off a body or a room, and a model trained to read what that reflection means. If you're curious about the broader range of ideas we're testing, our personal fun projects page and the full Uncategorized archive cover the rest.

Frequently asked questions

Does a WiFi wireless keyboard need any special hardware installed in the room?

The concept relies on existing WiFi infrastructure rather than dedicated sensors. Research systems like WiKey, as reported by EDN, used a standard laptop and a consumer router, while Macquarie University's FingerDraw used one WiFi transmitter and two receivers. The exact hardware needed for a finished product isn't something we've published specifics on yet.

How accurate is WiFi-based finger tracking compared to a physical keyboard?

Accuracy depends heavily on conditions. Macquarie University reports a median tracking accuracy of 1.27 cm for finger-drawing motion, and EDN reported WiKey reaching 97.5% keystroke accuracy under best-controlled conditions, falling to 77.5% once other people were moving in the same room.

Can other people in the room interfere with WiFi finger tracking?

Yes. EDN's reporting on WiKey shows accuracy falling from 97.5% in best-controlled conditions to 77.5% in a simulated real-world scenario with multiple people moving nearby. Any practical system has to separate the target's finger movements from background motion caused by other bodies in the same space.

Is WiFi finger tracking related to WiFi heart rate or pose sensing?

Yes, all three rely on the same underlying principle: a WiFi signal reflects off a body, and the reflection changes with movement. Our WiFi heart rate monitor concept reads chest wall movement from breathing, while pose and volumetric capture read larger body and room-scale reflections instead of finger-scale ones.

If you've got a specific use case in mind for wireless finger tracking, or you want to talk through how a WiFi keyboard concept might fit your own project, send your question through the enquiry form below and we'll get back to you directly.