A Recurrent Temporal Model for Semantic Levels Categorization Based on Human Visual System

Figure 8

Ten categories of images are used to train and test the model. The images of cars, motorcycles, helicopters, fishes, and dolphin images are all inside view; the images of cats and dogs are in front view; and airplanes, spiders, and butterflies are in top view. There are also, respectively, 10, 5, and 2 categories at the subordinate, basic, and superordinate levels in the dataset.