Thermometry of one-dimensional Bose gases with neural networks

TL;DR

This paper introduces a neural network approach for thermometry of one-dimensional Bose gases, achieving comparable accuracy to traditional methods but with fewer measurements, and providing insights into the features used for temperature prediction.

Contribution

The paper presents a neural network model that efficiently predicts temperature from absorption images of Bose gases, reducing the number of realizations needed compared to traditional density ripples thermometry.

Findings

Neural network predictions are compatible with established methods.

Fewer realizations are needed for similar precision.

Feature map analysis reveals relevant physical features.

Abstract

We design a neural network to extract and process features from absorption images taken of one-dimensional Bose gases in the quasi-condensate regime. Specifically, the network is trained to predict both the temperature of single realizations of the system and the uncertainty thereof. For multiple realizations, the individual predictions can be combined in an estimate of the mean temperature, improving precision. We benchmark our model on both simulated and experimentally measured data and compare it to the established method of density ripples thermometry. We find the predictions of the two methods compatible, although the neural network reaches similar precision needing much fewer realizations, thus highlighting the efficiency gain achievable when incorporating neural networks into analysis of data from cold gas experiments. Further, we study feature maps to reveal which local features of the condensate are extracted by the network and how said features correlate with properties of the system. A similar analysis could be employed to uncover physical relations in more complex systems.