Event-Based Visual Microphone.

We propose event-based visual microphone (EBVM), a passive electro-optical technique for capturing audio signals remotely using an event camera. The event-based camera records local deformations of a surface induced by the sound propagation by observing the changes in specular reflections at each pixel. Interpreting the timings of the specular incidences deduced from the event stream as signal level-crossings, we reconstruct the audio signal by imposing Fourier sparsity. The recovered audio signal is qualitatively comparable to or better than the prior art (intensity-based visual microphone), with three orders of magnitude lower data throughput.