Fig. 1: Calibration architecture and results for DCF-based system.
From: Multi-wavelength optical information processing with deep reinforcement learning
a depicts the calibration link. b presents a photograph of the encapsulated chip of the Kerr comb. c displays the 50 optical comb lines used in the system, which were generated from the Kerr optical comb. d shows the iterative cycles to the 4 calibration algorithms when using laser sources with different numbers of comb lines. e and f, respectively present the comparative outcomes of the DCF-based system output with and without DRC calibration, using 20 comb lines and 50 comb lines as laser sources
