Figure 6.

Visual representation of the classification accuracy results from Table 1. The number 512 at the end of some deep learning architecture names indicates a variant of the model that accepts 512 × 512 pixel images as input.