Skip to main content
. Author manuscript; available in PMC: 2021 Oct 28.
Published in final edited form as: IEEE Access. 2021;9:140824–140834. doi: 10.1109/access.2021.3118290

FIGURE 2.

FIGURE 2.

2-D U-Net architectures used in this study. (A) 2 stages, (B) 4 stages, and (C) 6 stages of encoder layers were implemented. Length of feature vectors was defined by n and varied between 32 to 256 as powers of 2. Increasing n improved the network capacity, with costs of increased memory requirements, and longer training and testing durations.