Doppler Gesture Sensing in JavaScript

[Daniel] stumbled on an interesting paper (which we featured before) on Doppler gesture sensing using only a computer’s speaker and microphone. Unfortunately the paper didn’t include source code so [Daniel] created his own implementation of Doppler gesture sensing in JavaScript that works right in the browser.

[Daniel]’s JavaScript library generates a sine wave at 20 kHz that’s played through the computer’s speakers. The frequency is high enough that it’s pretty much inaudible. While the tone is being played through the speakers, the computer’s microphone is used to sample the audio and calculate the frequency spectrum of the signal. As you move your hand closer to the computer while the tone is playing, the frequency of the received signal shifts higher; as you move your hand away, it shifts lower. [Daniel]’s script looks for this frequency shift and uses it to trigger events.

doppler

[Daniel] has some awesome examples included on his website where you can test out the functionality for yourself. He has a hands-free scrolling example, spectrum plot, and even a virtual theremin. Since his code is bundled up into an easy-to-use library, it should be fairly easy to integrate into any webpage. The only real limitation to the library is that it only works in Chrome right now (Firefox doesn’t support disabling echo cancellation).

DIY Hearing Aid

DIY Hearing Aid

Hearing aids are expensive little devices, typically costing a few thousand dollars each. They need to be highly integrated to fit in the ear, while still providing signal processing to ensure good audio quality.

This DIY hearing aid does some intelligent signal processing. It uses an electret to capture audio, then uses a pre-amplifier to increase the gain 100 times. The next stage consists of four filters, dividing the input signal by frequency into four parts. These are passed into four LTC6910 programmable gain amplifiers, which allow an Arduino to control the gain of each channel. The LTC6910 takes 3 digital inputs that are used to set the gain value.

To determine which gain to use for each frequency band, the Arduino needs to know how much power is in each band. This could be done using a Fast Fourier Transform, but that would require quite a bit of processing power. Instead, an envelope detector averages the signal, which can be read by an analog input on the Arduino. Using this information, the hearing aid can boost specific frequencies when it detects conversation.

This hearing aid won’t quite fit in your ear, but there is a lot of interesting signal processing going on. The schematic, Arduino source code, and a MATLAB simulation are provided.