CH. 3 CP
Introduction
Perception: Experiences resulting from stimulation of the senses.
Key Questions
Why can two people experience different perceptions in response to the same stimulus?
How does perception depend on a person's knowledge about characteristics of the environment?
How does the brain become tuned to respond best to things likely to appear in the environment?
What is the connection between perception and action?
Crystal's Beach Experience
Crystal's run along the beach illustrates how perception can change based on added information and involve reasoning or problem solving.
Initial Perception: Mistakenly sees an umbrella as driftwood in misty conditions.
Correction of Perception: Realizes it is an umbrella upon closer inspection.
Nature of Perception
Perception is not automatic; it involves complex processes.
Example of coiled rope: Assumed to be continuous based on past experience.
Perception is linked with action—Crystal’s running and later grasping her coffee cup demonstrates coordination between seeing and physical interaction.
Characteristics of Perception
Dynamic Process: Rapid adjustments based on knowledge and sensory input.
Reasoning Involvement: Sometimes mimics reasoning or problem solving.
Interaction with Action: Essential for tasks such as grasping objects, which involve continuous spatial awareness.
Understanding Scene Perception
Example: Roger viewing Pittsburgh from PNC Park demonstrates the ability to interpret various buildings and objects in a scene.
Perceptual Puzzles: Humans can interpret complex visual scenes instinctively, while computers struggle to achieve the same understanding.
Computer vs. Human Perception
Computers have historically struggled with visual perception despite advancements.
Early systems required excessive processing time versus rapid human interpretation.
Challenges for Computer Vision:
Ambiguous stimulus on receptors (inverse projection problem).
Objects potentially hidden or blurred.
Objects appearing different from alternate viewpoints.
High-level contextual information needed for scene understanding.
Mechanisms of Perception
Bottom-Up Processing: Starts with stimulation of the receptors leading to image transmission to the brain.
Top-Down Processing: Involves existing knowledge impacting perception, aiding recognition and meaning-making in context.
Examples:
Blobs perceived differently in varying contexts based on prior knowledge.
Rapid object identification in scenes, informed by statistical learning acquired over time.
Approaches to Object Perception
Helmholtz's Theory
Unconscious Inference: Perceptions are based on past experiences (likelihood principle).
Gestalt Principles
Emphasize innate organizing principles over learned experiences.
Illustrate principles such as good continuation and simplicity in perception.
Regularities in the Environment
Physical Regularities: Frequent features in human-made and natural environments influence perception.
Semantic Regularities: Relate to the inferred meanings and typical functionalities within scenes.
Bayesian Inference
Combines prior probabilities with likelihood based on new evidence to enhance perceptual conclusions.
Neural Basis of Perception
Experience-Dependent Plasticity: Neurons adapt to respond to common environmental characteristics.
Neuroscientific Evidence: Specialized neurons respond primarily to environmental regularities, established through evolutionary adaptation.
Action and Perception Interaction
Movement as a Tool for Perception: Movement allows for enhanced understanding through revealing hidden aspects of objects.
Coordination in Actions: Everyday actions, such as reaching for a coffee cup, require dynamic integration of perception and action, facilitated by separate but coordinated neural pathways.
What and Where Streams
Perception Pathway (What): Identifies objects.
Action Pathway (Where): Facilitates spatial location for interactions.
Mirror Neurons
Mirror neurons fire during both action execution and observation, linking perception and action more closely.
They may offer insights into the intentions behind observed actions, guiding social interactions and understanding.
Conclusion
Perception is complex, integrating sensory input, prior knowledge, and action capabilities into a coherent interaction with the environment.