Erlangen 2022 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
A: Fachverband Atomphysik
A 6: Interaction with strong or short laser pulses II
A 6.2: Vortrag
Dienstag, 15. März 2022, 11:00–11:15, A-H1
Retrieval of the internuclear distance in a molecule from photoelectron momentum distributions using convolutional neural networks — •Nikolay Shvetsov-Shilovski and Manfred Lein — Leibniz Universität Hannover, Hannover, Germany
We train and use a convolutional neural network (CNN) to recognize the internuclear distance of a two-dimensional H2+ molecule from the photoelectron momentum distribution produced by a strong few-cycle laser pulse [1]. We show that the CNN trained on a dataset consisting of a few thousand images can retrieve the internuclear distance with the mean absolute error less than 0.1 a.u.
We investigate the effect of the focal averaging on the retrieval of the internuclear distance. The CNN trained on a set of focal averaged momentum distributions also shows good performance in recognizing of the internuclear distance: the corresponding mean absolute error does not exceed 0.2 a.u. Furthermore, we compare the application of the CNN with an alternative approach based on the direct comparison of the momentum distributions.
[1] N. I. Shvetsov-Shilovski and M. Lein, submitted to Phys. Rev.A, arXiv:2108.08057.