What categories of images are documentary and what kinds are not?
Among the tasks that computers are bad at is recognising broad categories of images.
There are three categories of images, representing the peaceful, moderate, and angry deities, and three schools of painting, the Sman-thang, Gong-dkar Mkhan-bris, and Kar-ma sgar-bris, which are differentiated by colour tones and depicted facial expressions.
More specifically, let w the number of categories of images in the training set.
Looking into the various categories of images, EyeEm Vision was the most accurate for cityscapes, people/sports, nature and animals (all at 83%).
Experiments show that setting T to be 5-10% quantile of the histogram for | ∇ u σ | 2 is sufficient for wide categories of images.
The different categories of images were scrambled and presented separately on a computer screen using Microsoft PowerPoint.
Given 9 categories of images, and 9 pictures per category, pictures were presented in a counterbalanced order such that no condition followed or preceded another more than once.
Currently, we are looking for common second-order properties of different categories of images in general and restrict our investigation to the power spectrum.
In order to determine whether the Fourier power spectra of man-made images differed from those of other categories of images (photographs of natural scenes, objects, faces and plants and scientific illustrations), we analyzed their 2D power spectra by principal component analysis.
