Spike-based signal processing

Traditional sampling focuses on encoding and decoding signals by recoding the amplitude values of the signal at known time points. Time encoding is an alternative approach to classical sampling, inspired by the processing of information of spiking neurons in the human brain.

In this talk we consider time encoding of some classes of non-bandlimited signals, such as streams of Diracs and piecewise constant signals, and show how they can be fully recovered from their timing information.

We provide sufficient conditions for perfect recovery of the signals and present an algorithm to perform the reconstruction, as well as some simulation results. Finally, we suggest some open research questions in the area of event-based sensing and processing.