perceiving object, Motions, color, Hearing

• Object recognition


  Involves analyzing an image to detect and classify objects present in it. It relies on feature extraction, object detection, classification, and recognition to accomplish this task.
• The computer recognized the object as: (1) a boy, (2) a baseball bat, (3) then created a description of the scene.
• Why it is difficult to design a perceiving machine

  • The stimulus on the receptors is ambiguous
  • Objects can be hidden or blurred
  • Objects look different from different viewpoints
• Inverse projection problem


  Involves interpreting the retinal image and extending imaginary lines out from your eye to determine the possible objects that could have produced it

  
• The retinal image could also be created by other objects, like a tilted trapezoid or a larger rectangle, depending on their position and orientation relative to your eye

  
• The two-dimensional image on your retina may not always accurately represent what's present in the environment, highlighting the complexities of perception and interpretation

  
• When objects are hidden or blurry, people are surprisingly good at finding them
• Despite the fuzzy images, people can often identify most of them. But computers struggle more with this kind of task, as they're not as good at interpreting blurry or obscured images
• Viewpoint invariance


  • The ability to recognize objects from different viewpoints, which is tough for computers to do
• Perceptual organization


  How our brains make sense of what we see by grouping and separating visual elements to form coherent perceptions

  
• Grouping


  When your brain puts together similar visual elements to create objects
• Segregation


  When your brain separates one object from another
• Structuralism


  Viewed perceptions as complex experiences built from simpler sensations
• Gestalt psychology


  Emphasized how our brains organize and interpret sensory information as meaningful wholes
• Apparent movement


  Phenomenon where our brain perceives movement even though nothing is physically moving

  
• Apparent movement challenges the idea that perceptions are merely constructed by combining individual sensations
• Illusory contours


  Edges that our brain perceives but aren't physically there

  
• Gestalt principles of perceptual organization

  • Good continuation
  • Pragnanz
  • Similarity
  • Proximity
  • Common fate
  • Common region
  • Uniform connectedness
• Good continuation


  Our brain tends to perceive continuous lines or patterns rather than broken or disjointed ones
• Pragnanz


  Our brains tend to interpret stimuli in the simplest and most straightforward way possible
• Similarity


  Similar things tend to be grouped together
• Proximity


  Things that are close to each other appear to be grouped together
• Common fate

  Things moving in the same direction seem to be grouped together
• Common region


  Elements within the same spatial region appear grouped together
• Uniform connectedness


  Connected regions of the same visual properties, like lightness or color, are perceived as a single unit
• Our assumptions based on past experiences help our brains organize what we see
• Perceptual segregation


  How we perceive objects as separate from their background
• Properties of figure and ground

  • Memorability
  • Position
  • Border
• Border ownership


  The border separating the figure from the ground seems to belong to the figure
• Properties of the image that determine which area is figure

  • Areas lower down in the picture are more likely to be seen as the main figure
  • Borders that curve outward (convex) make the part inside the curve more likely to be seen as the figure
• Our past experiences don't always determine how we perceive things, as certain built-in rules or principles can be more important
• Recognition by components (RBC) theory


  Objects are made up of basic shapes called "geons" and most objects can be recognized based on how these geons are arranged
• RBC theory explains why we can recognize objects from different angles, but it has limits in explaining how we group objects together and doesn't account for differences between similar objects
• Perceiving scenes and objects in scenes

  Scenes are views of real-world environments that include background elements and multiple objects organized in relation to each other
• Gist of the scene


  The quick understanding of what's happening in a scene
• People can identify a scene after seeing it for just a quarter of a second, even when it's described in words
• Global image features that help recognize the gist of a scene

  • Degree of naturalness
  • Degree of openness
  • Degree of roughness
• scene, like a car chase or a quiz show, even though we don't focus on every detail
• Mary Potter found that people could identify a scene after seeing it for just a quarter of a second, even when it's described in words
• Li Fei-Fei showed that people could understand scenes in as little as 27 milliseconds