Skip to main content
. 2021 Sep 9;11(9):2655. doi: 10.3390/ani11092655

Figure 4.

Figure 4

Example images that were annotated by the neural network. On the left side, the network was trained with 3400 images labeled with naive annotations of human observers; on the right side, it was trained with 3400 images labeled by network-assisted annotations obtained from human observers. The human annotations are framed with green bounding boxes and those of the network are framed with red bounding boxes.