“…We perform this sectional estimation capitalizing on a convolutional neural network (LeCun et al, 1998). The CNN has recently been identified as one of the promising tools for data-driven fluid flow analyses including state estimation (Fukami et al, 2019a(Fukami et al, , 2021aKobayashi et al, 2021), flow control (Lee et al, 1997;Rabault et al, 2019), reduced-order modeling (Murata et al, 2020;Hasegawa et al, 2020b,a;Fukami et al, 2020c;Kim and Lee, 2020;Nakamura et al, 2021b;Maulik et al, 2021), and turbulence modeling (Fukami et al, 2019b;Duraisamy et al, 2019;Lapeyre et al, 2019;Pawar et al, 2020;Thuerey et al, 2020;Font et al, 2021), thanks to a filter operation inside the CNN (Fukami et al, 2020a).…”