Perception

The Nature of Perception

  • Perception is defined as experiences resulting from the stimulation of the senses.
  • Visual perception is a dynamic process influenced by more than just receptor stimulation.
  • Perceptions can change with added information and involve reasoning or problem-solving.
  • Perception can be based on perceptual rules derived from past experiences.
  • Arriving at a perception can be a process that takes time, resembling reasoning.
  • Perception occurs in conjunction with action, supporting dynamic processes.

Going Beyond Light-Dark Patterns

  • Perception involves easily distinguishing different buildings and objects in a scene.
  • It requires going beyond the pattern of light and dark on the retina.
  • Computer vision systems still struggle to match human perceptual recognition.

Computer Vision vs. Human Perception

  • Computer vision systems are actively researched for fine-grained judgments.
  • They still perform below human capabilities in tasks like recognizing species.
  • Current computer programs often make errors humans would never make.
  • Computer vision programs can be easily fooled by white noise static patterns.
  • Humans outperform computers in recognizing faces at angles or with changes.

Why is it so Difficult to Design a Perceiving Machine?

  • Humans solve perceptual problems easily, quickly, and automatically.
  • The stimulus on the receptors is ambiguous.
  • The image on the retina can be created by various objects at different distances (inverse projection problem).

Ambiguities in Perception

  • Objects can be hidden or blurred, requiring effort to locate them.
  • People use knowledge of the environment to determine what is likely present.
  • Objects look different from different viewpoints (viewpoint invariance).

Information for Human Perception

  • Perception relies on information from the environment, such as light for visual perception.
  • The sequence from eye to brain involves bottom-up processing.
  • Bottom-up processing starts with environmental energy stimulating receptors.
Bottom-Up Processing
  • Receptors translate energy into neural signals transmitted to cortical brain areas.
  • Perception involves more than just receptor activation and bottom-up processing.
  • Top-down processing originates in the brain, involving knowledge and expectations.
Top-Down Processing
  • Perception is a combination of both top-down and bottom-up processing.
  • Top-down processing is influenced by:
    • Knowledge
    • Expectations
    • Experience
    • Memories
    • Culture

Examples of Top-Down Processing

  • Top-down processing examples include:
    • Perceiving visual objects and people
    • Hearing words in a sentence
    • Experiencing pain
Perceiving Objects and People: The Role of Context
  • Context influences perception; a central stimulus can be seen as a letter or number based on its surroundings.
  • Context affects how we identify people and assess their emotional expression.
  • People use body information to make identification judgments when facial features are ambiguous.
  • Emotional context can affect how participants rate the emotional expression of a neutral face.
Hearing Words in a Sentence: The Role of Knowledge and Experience
  • Knowledge and experience influence perception of speech.
  • Listeners familiar with a language perceive individual words in a continuous sound stream (speech segmentation).
  • Continuous sound signal enters the ears (bottom-up), and language knowledge (top-down) creates individual word perception.
Experiencing Pain: The Influence of Attention
  • The perception of pain is influenced by top-down processing.
  • The direct pathway model suggests pain occurs when nociceptors are stimulated and send signals to the brain (bottom-up).
  • Pain perception can be influenced by expectation, attention, and distracting stimuli.
  • Patients who are informed and relaxed request fewer painkillers.
  • Placebos can provide real pain relief due to expectation.
  • Distraction can reduce pain behavior and subjective distress.
  • Perception is created by both signals from the environment (bottom-up) and what the individual brings to it (top-down).

Conceptions of Object Perception

Helmholtz's Theory of Unconscious Inference
  • Hermann von Helmholtz proposed the theory of unconscious inference.
  • Ambiguity means a pattern on the retina can be caused by various objects.
  • The likelihood principle states that we perceive the object that is most likely to have caused the stimuli.
  • Unconscious inference involves unconscious assumptions about the environment.
  • Perception resembles solving a problem, using knowledge to infer what the object might be.
  • This process happens rapidly and unconsciously.
The Gestalt Principles of Organization
  • The Gestalt approach originated as a reaction to structuralism and the belief that perceptions are formed by simply adding up sensations.
  • Gestalt psychologists proposed principles of perceptual organization.
  • The whole is more than the sum of its parts.
Figure-Ground Principle
  • The figure-ground principle is the organization of the visual field into objects (the figures) that stand out from their surroundings (the ground).
Good Continuation
  • The principle of good continuation states that points forming straight or smooth curves are seen as belonging together.
  • Overlapping objects are perceived as continuing behind the overlapping object.
Pragnanz
  • Pragnanz, or "good figure," states that perceptual fields take on the simplest and most encompassing structure.
Similarity

  • The principle of similarity states that similar things appear to be grouped together by:

    • color
    • size
    • shape
    • orientation
    • enclosure
    • proximity

  • Wertheimer described these principles as "intrinsic laws" built into the system.

  • Experience plays a minor role compared to perceptual principles, differing from Helmholtz's likelihood principle.

Taking Regularities of the Environment into Account
  • Modern cognitive psychologists suggest perception is influenced by environmental regularities.
Physical Regularities
  • Physical regularities include the prevalence of vertical and horizontal orientations.
  • The light-from-above assumption affects perception of protrusions and recesses.
Semantic Regularities
  • Semantic regularities relate to the meaning of a scene and its associated functions.
  • Scene schema is knowledge of what a given scene typically contains.
  • Context influences perception, as demonstrated by experiments where objects appropriate to a scene are more easily identified.
Bayesian Inference
  • Bayesian inference uses prior probability and likelihood to estimate the probability of an outcome.
  • Posterior∝Prior⋅LikelihoodPosterior \propto Prior \cdot Likelihood
  • Prior probability is our initial belief, and likelihood is the extent to which evidence is consistent with the outcome.
  • People start with a prior probability, use additional evidence to update their belief, and reach a conclusion.
  • Applying this to object perception, the inverse projection problem is addressed by combining retinal images with prior probabilities based on past experiences.
  • Bayesian inference reformulates Helmholtz's idea in terms of probabilities.
Comparing the Four Approaches
  • Helmholtz, regularities, and Bayesian inference use data about the environment.
  • Gestalt psychologists emphasize built-in principles, though modern psychologists suggest these could be shaped by experience.

Neurons and Knowledge About The Environment

  • Neurons in the visual cortex respond to horizontals and verticals more than obliques. An effect known as The oblique effect
  • Evolution and experience-dependent plasticity shape the response properties of neurons.
Experience-Dependent Plasticity
  • The brain is changed by exposure to the environment, improving perceptual efficiency.
  • Neurons become tuned to specific aspects of the environment.
  • Studies on kittens raised in environments with only vertical stripes showed their brains developed neurons that responded mainly to verticals.
Experience-dependent plasticity in humans

  • An area in the temporal lobe called the fusiform face area (FFA) contains many neurons that respond best to faces.

  • Experience-dependent plasticity has been demonstrated in humans using fMRI.

  • Training can cause neurons to respond to complex objects other than faces.

  • Neurons in the FFA respond strongly to faces due to both nature and nurture.

The Interaction Between Perceiving and Taking Action

  • Perception typically occurs in dynamic situations with movement and action.
Movement Facilitates Perception
  • Movement reveals aspects of objects not apparent from a single viewpoint.
The Interaction of Perception and Action
  • Perceiving and acting are coordinated continually.
  • Picking up a cup of coffee requires identifying the cup, positioning fingers, and applying the right amount of force.
The Physiology of Perception and Action
  • There are two processing streams in the brain:
    • One for perceiving objects (what pathway)
    • One for locating and acting toward objects (where pathway).
  • Brain lesioning and neuropsychology reveal principles about the normal brain.
What and Where Streams
  • Ungerleider and Mishkin's experiment showed that:
    • Damage to the temporal lobe affects object discrimination (what pathway)
    • Damage to the parietal lobe affects landmark discrimination (where pathway).
Perception and Action Streams
  • Milner and Goodale (1995) used neuropsychological approach (studying the behaviour of people with brain damage) to reveal two streams:
    • Temporal lobe: (ventral stream)
    • Parietal lobe: (dorsal stream)
  • Neuropsychological studies reveal two streams, one involving the temporal lobe and the other the parietal lobe.
  • D.F., who suffered temporal lobe damage, could not recognize objects but could guide her hand movements.
  • D.F. performed poorly on static orientation matching tasks but well when action was involved.
  • This suggests separate mechanisms for judging orientation and coordinating vision and action.

Pioneering Studies: Attention as Selection

  • Early research on attention helped establish the information processing approach to cognition.
Broadbent's Filter Model of Attention
  • Attention became an important research after the world war 2.
  • Broadbent's filter model of attention (Broadbent, 1958)
  • dichotic listening (presenting different stimuli to the left and right ears)

Spatial Attention: Overt and Covert Attention

Overt Attention
  • Cognitive factors based on knowledge of the environment. Example (stop signs positioned at junctions and those positioned in the middle of a street)
  • Scanning based on task demands.
Covert Attention
  • Directing attention without eye movements\, a process called covert attention
  • Covert attention is an important part of many sports.