Invited Talk: Cristian BonatoBayesian inference for quantum sensing and model learning

The development of techniques for the characterization of quantum states and their dynamics is crucial for applications in quantum communication, computing, sensing and simulation. In quantum sensing, for example, physical parameters of interest can be measured, with quantum-limited sensitivity and high spatial resolution, by optimizing information extraction from a quantum sensor.

Here we present our progress on the application of Bayesian inference for the characterization of quantum systems. In a first set of experiments, we apply sequential Bayesian estimation to improve the performance of a quantum sensor based on a single electronic spin. Our protocols, implemented on a fast micro-controller, adaptively choose optimal experimental settings in real-time to estimate decoherence timescales and static magnetic fields. We show that online adaptive approaches can provide a speed-up by a factor up to one order of magnitude compared to their non-adaptive counterparts, and discuss the expected impact on quantum sensing.

In a second set of experiments, we focus on automated reconstruction of the model for an unknown system of quantum emitters based on the arrival times of the emitted photons. We employ a Markov-chain Monte Carlo (MCMC) approach to infer the distribution of multiple parameters. We also introduce an algorithm to reconstruct a Lindblad master equation that describes the system, only setting an assumption on the number of energy levels involved. We benchmark our algorithms on a system of two resonant self-assembled InGaAs quantum dots in the cooperative emission regime.