Quantum computational imaging and sensing

TL;DR

Quantum computational imaging and sensing (QCIS) proposes to encode electromagnetic-field quantum information into a quantum computer for optimized, generalized measurements, enabling potentially exponential advantages in imaging and sensing under weak-field conditions. The approach combines quantum transduction (e.g., optomechanical interfaces) with variational quantum circuits to implement joint measurements that surpass classical receivers, even with noisy intermediate-scale quantum devices. The paper provides a concrete JDR example for coherent optical communication, showing potential improvements in the error probability $p_{ m err}$ in the regime $|\alpha|\ll 1$ given realistic transduction fidelities, and discusses the hardware and theory work needed to realize these gains. Overall, QCIS offers a pathway to practical quantum-enhanced imaging and sensing that could reduce light and data requirements for certain tasks, leveraging near-term quantum technologies.

Abstract

We present a new framework for imaging and sensing based on utilizing a quantum computer to coherently process quantum information in an electromagnetic field. We describe the framework, its potential to provide improvements in imaging and sensing performance and present an example application, the design of coherent receivers for optical communication. Finally, we go over the improvements in quantum technologies required to fully realize quantum computational imaging and sensing.

Figures (2)

• Figure 1: (a) Conventional imaging and sensing, where quantum information in the electromagnetic field is immediately transformed to classical information by intensity detection at the image plane. (b) Schematic of the quantum computation imaging and sending (QCIS) concept, where relevant quantum information is transduced into the state of a register of qubits, on which a quantum computation is performed before conversion to classical information.
• Figure 2: $p_{\rm err}$ in decoding two bits encoded in three optical pulses using binary phase shift keying (BPSK), as a function of the magnitude of the received optical pulses. The black curve shows the optimal performance of the optimal "classical" single pulse receiver that operates in the optical domain (and thus decodes two pulses, which is sufficient to encode 2 bits using BPSK). The blue (orange) dots show the $p_{\rm err}$ achieved by transducing to qubit states and performing an optimized variational circuit to discriminate between the possible codeword states, when the optomechanical system performing the transduction is at 1mK (1K).