Die DPG-Frühjahrstagung in Hannover musste abgesagt werden! Lesen Sie mehr ...
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
Q: Fachverband Quantenoptik und Photonik
Q 22: Posters: Quantum Optics and Photonics II
Q 22.12: Poster
Dienstag, 10. März 2020, 16:30–18:30, Empore Lichthof
Deep learning phase reconstruction from dispersion scan traces — •Sven Kleinert1,2, Ayhan Tajalli1,2, Tamas Nagy3, and Uwe Morgner1,2,4 — 1Institute of Quantum Optics, Leibniz Universität Hannover, Welfengarten 1, 30167 Hannover, Germany — 2Cluster of Excellence PhoenixD (Photonics, Optics, and Engineering Innovation Across Disciplines), Hannover, Germany — 3Max Born Institute for Nonlinear Optics and Short Pulse Spectroscopy, Max-Born-Straße 2a, 12489 Berlin, Germany — 4Laser Zentrum Hannover e.V., Hollerithallee 8, 30419 Hannover, Germany
For applications of femtosecond laser pulses the knowledge of their temporal shape plays a crucial role. Many different techniques for measuring these information have been developed during the last decades. The recently proposed dispersion scan (d-scan) is a simple technique to fully characterize ultrashort pulses. The reconstruction of the temporal shape of the pulses from a d-scan trace is an inverse problem as it is usually the case for the characterization techniques, which benefit from an inherent redundancy in the measurement. Conventionally, time-consuming optimization algorithms are used to solve these inverse problems. Here, we show the retrieval of femtosecond pulses from d-scan traces using a deep neural network. This neural network is trained with artificial and noisy dispersion scan traces from randomly shaped pulses. After the training, the phase reconstruction takes only 16 ms enabling video-rate pulse characterization. This retrieval shows a great tolerance against noisy conditions, delivering reliable retrievals from traces with up to 20% of noise.