As the fully-connected layers are horizontal, the images, which in general are three-dimensional structures (width × height × channels), must be flattened and transformed into one-dimensional arrays where the geometric properties are definitively lost. With more complex datasets, where the distinction between classes depends on more details and on their relationships, this approach can yield moderate accuracies, but it can never reach the precision required by production-ready applications.
The conjunction of neuroscientific studies and image processing techniques suggested experimenting with neural networks where the first layers work with bidimensional structures (without the channels), trying to extract a hierarchy of features that are strictly dependent on the geometric properties of the image. In fact, as confirmed by neuroscientific research about the visual cortex, a human being doesn't decode an image directly. The process is sequential and starts by detecting...