Perception is our sensory experience of the world. Sensation entails the processes by which physical properties are converted to neural signals. Perception does not give us a faithful image of the real-world. There are three fundamental concepts of perception:

1. Inverse problem
   The world is three-dimensional, but the image projected on our fovea is two-dimensional. This shows that our brain has lost a bit of information in the perception (the fact that the object is three-dimensional) and our brain has to make up for that when inverting the image back to a three-dimensional image.
2. Bottom-up and top-down processing
   In bottom-up processing the original sensory input into a percept. It is data-driven. Top-down processing occurs when the way we perceive things changes the way we see things. Bottom-up processing probably dominated with unambiguous information and top-down processing dominates with ambiguous information.
3. Likelihood principle
   The likelihood that an object or event will occur is important for the perceptual processing of that object or event (e.g: if we are looking for cats we are more likely to perceive things as cats than otherwise).

Invariants in vision are properties of the three-dimensional object being viewed that can be derived from any two-dimensional image of the object. Direct perception refers to the bottom-up process by which objects and their function are recognized. Direct perception consists of three levels:

1. Computational theory
   In this level, perception focusses on what the purpose of computation is and why it does what it does (bottom-up)
2. Choice of representation
   There is a transformation between input and output (top-down)
3. Achieving the computations
   How perception actually takes place. This is biologically limited.

Embodied cognition holds that cognition is about the experiences arising from a perceptual system tightly linked to an action system, rather than the manipulations of abstract representations.

Synaesthesia is a condition where stimulation of one perceptual modality results in experiencing a percept in a typically unrelated modality (e.g: seeing letters in different colours). The basic organization of a perceptual system is a hierarchy (e.g: eyes – cones – optic nerve, thalamus, cortex – visual cortex – colour). Somatoperception is the perception of touch and proprioception is the sense of how our limbs are positioned in space and vestibular sensation is the sense of balance and orientation in space.

The encoding of visual information begins in the retinas of the two eyes and is transmitted from there to the primary visual cortex. Receptors in the eyes transform light into a neural signal. The image is projected on the fovea. The eye has two types of receptors:

1. Cones
   These encode colour and high-resolution spatial form information. They are mostly centred in the fovea and there are fewer cones the further away you move from the fovea.
2. Rods
   These encode motion and low-resolution form information. They cannot see colour. They are mostly around the fovea.

The right visual field ends up in the left hemisphere of the brain and the left visual field ends up in the right hemisphere of the brain. The place where the neurons come together in the fovea is a blind spot and the brain fills in this information by itself. The fovea has a disproportionate amount of visual cortex dedicated to processing incoming visual information, thus explaining why we focus on the centre of our sight. Things that are viewed close together, are also processed close together. There are two primary pathways for visual processing that lead into the occipital cortex and beyond:

1. Ventral stream
   This pathway leads from the visual cortex to the temporal lobe and is specialized in object determination. It is also called the “what-pathway”.
2. Dorsal stream
   This pathway leads from the visual cortex to the parietal cortex and is specialized for determining where objects are in the visual world. It is also called the “where-and-how pathway”. This pathway is crucial for guiding movements when picking up objects.

The experiment by Hubel and Wiesel shows that there are feature detectors in vision. Those are neurons that fire when seeing a certain feature (e.g: neurons firing when a bar is vertical or almost vertical and not firing when the bar is horizontal).

The encoding of auditory information begins within a special structure in the ear known as the cochlea and is transmitted from there to the primary auditory cortex. On the cochlea, there is a band of nervous tissue called the basilar membrane and on this membrane are hair cells that move in response to sound pressure to transform it into a neural signal. The auditory system is arranged tonotopically. The frequency and thus the pitch of a sound is determined by the amplitude. The loudness is determined by the firing rate of the neurons. Aphasia is the inability to use either verbal or written language and amusia is the inability to hear differences in tones and is also called tone-deafness. Phonagnosia is the inability to recognize familiar sounds, although they can discriminate between sounds.

The processing of touch begins in specialized receptors in the skin, which project pathways of neurons to the brain. These pathways end in the somatosensory cortex which is located next to the central sulcus, the major boundary between parietal cortex and frontal cortex. Sensitiveness of touch is not equally divided over the body. Brodmann areas are regions in the brain distinguished by the structural properties of the neuronal architecture.

Information about the world should be combined because each piece of information has its own strengths and weaknesses. All the systems are linked together in forming perceptions. The McGurk effect demonstrates that the combination of sensory information can lead to a perception that is different from that produced by individual sources (e.g: we see something different than we hear and thus we hear something different). Intelligibility increases when sight and sound are put together. The modality appropriate hypothesis states that for each physical property of the environment, there is a particular sensory modality that has a higher acuity for estimating this property than the other senses. This modality will always dominate bimodal estimates of the property (e.g: the sense that is most useful in a task dominates). Vision dominates most of the time.

General perceptual processes produce an object representation that can be compared to a stored internal representation and thus we recognize things. Feature analysis states that we recognize objects because we recognize the components that are included in our ‘internal library’ and put them together. There are some problems with this theory because it requires us to have an internal image of all possible scenarios for all possible objects and this would require immense memory storage.

The pandemonium model states that there are so-called demons in a hierarchy, with lower demons assigned to evaluate the utility of individual features and higher-level demons assigned with evaluating the success of lower-level demons (e.g: when reading a sentence, one set of demons recognizes lines, one set of demons recognizes the letters in those lines and so on). The prototype theory states that we recognize objects because we compare the object we’re seeing to the ‘best / most common’ version of that object that we know.

Viewpoint invariant relationship is an aspect of an object that we view that is preserved, no matter the direction from which we view the object. Geons are the elements of a set of volumetric primitives or shapes that can be recognized from any viewpoint. Geons are crucial In Biederman’s recognition by component theory. The multiple views theory states that recognition is image-based and that object recognition is achieved by storing representations of a few select views of the object that had been learned. This theory predicts that we are better at recognizing objects if we look at them from directions more similar to the learned views. The information theory states that patches of images with intermediate complexity were optimal to encode a set of images for subsequent recognition tasks. In human recognition using the view, colour is only important when form is ambiguous.

Haptic perception is the combination of abilities that allow us to represent the material characteristics of objects and surfaces for recognition (e.g: we use our hands to discover what we are holding, somatoperception system). People with visual agnosia are unable to recognize objects. People with prosopagnosia can’t recognize faces anymore.

Event perception refers to the changes in layout, changes in surface existence or changes in colour and texture or it refers to an event as a segment of time at a given location that is conceived by an observer to have a beginning and end. The perceptual cycle is defined as the cyclic process comprised of the following steps:

1. Memory in the form of schema drives exploration
2. This leads to information pick-up
3. This leads to potential modification of schema and subsequent repetition of the steps of this cycle

A schema is a framework that represents a plan or theory, supporting the organization of knowledge.

Social perception informs us about the thoughts, emotions and internal states of others and this is useful information to help us navigate our social world. People with Capgras syndrome believe that people or things have been replaced with duplicates. There are some general properties for face recognition:

1. Familiar faces
   Humans are especially tuned to recognize familiar faces.
2. Unfamiliar faces
   Humans are poor at recognizing unfamiliar faces.
3. Brain areas
   Humans have specialized brain areas and networks for facial recognition.
4. Holistic
   The mechanisms of facial recognition are holistic.

The cross-race effect refers to the fact that people are more likely to identify someone from the same race than someone from another race. Haxby’s neural model of human face recognition comprises multiple regions spread throughout the brain. The organization of this system emphasizes a distinction between the representation of invariant and changeable aspects of the face. The invariant aspects of faces are responsible for the recognition of individuals and the changeable aspects of faces facilitate social communication.

One way voice carries information, independent of linguistic content, is found in the fact that the emotional content of an utterance can be carried in the prosody of the speech. The prosody is the rhythm, intonation and stress patterns in speech. People with phonagnosia can’t recognize identity from voices. Humans are able to recognize identity by looking at the biological motion. This way we can determine action, identity, vulnerability and sometimes gender.