Spike-based sampling is an alternative approach to classical (Shannon-based) sampling. For this sampling scheme, data is only acquired after a signal-dependent event (e.g. when the amplitude of a signal changes by a certain amount). After such an event a spike is triggered. This e.g. allows for a more efficient data encoding compared to classical sampling. Spike-based sampled signals require different learning algorithms than conventionally sampled signals. Examples of such learning methods are Spiking Neural Networks (SNN). This project jointly investigates spike-based sampling and learning. It aims at developing new spike-based sampling schemes and novel spike-based learning algorithms. It will cover the whole range from the mathematical foundation to prototype implementation demonstrating the capabilities of spike-based sampling and learning.