Prof. Greg Francis1PSY 310: Sensory and Perceptual ProcessesPurdue UniversityObject featuresPSY 310Greg FrancisLecture 16What’s odd here?Purdue UniversityPerceptual organization The Gestalt laws of perceptual organization Pragnanz (good figure) Similarity Good continuation Proximity Common Fate Familiarity And other similar kinds of laws Are all about putting together information in a scene to identifyobjects An important part of this is figure-ground separationPurdue UniversityFigure-ground Objects occlude other objectsPurdue UniversityFigure-ground How do you know the edge of the tree belongs to the tree?Purdue UniversityFigure-ground Part of the explanation is that other contours seemto go behind the treePurdue UniversityFigure-ground The common color of the house and archway helps connectthose parts of the image Which implies that the tree is in frontProf. Greg Francis2PSY 310: Sensory and Perceptual ProcessesPurdue UniversityContours An important part of this process is identify to which side anedge “belongs” Here, the white-to-black edge belongs to the white squarePurdue UniversityContours An important part of this process is identify to which side an edge“belongs” Here, the black edge belongs to the black half-moonPurdue UniversityContours The black side of the edge is the same in both casesPurdue UniversityBackground Our perception of what we experience in the backgrounddepends on the foreground You know this is a checkerboardPurdue UniversityBackground Yet, we sense that the color of the region behind the gray discs is thesame as the surrounding checksPurdue UniversityGrouping Identifying thispattern ofelements isdifficult Why?Prof. Greg Francis3PSY 310: Sensory and Perceptual ProcessesPurdue UniversityGrouping It’s easier with anoccluder Why?Purdue UniversityGrouping Camouflage is basically working on this principlePurdue UniversityChoices Figure-ground is significant because it seems thevisual system makes a “choice” about things You are not aware of making these choices, it’s not adecision in the traditional sensePurdue UniversityChoices Figure-ground is significant because itseems the visual system makes a“choice” about things You are not aware of making thesechoices, it’s not a decision in thetraditional sense If you see one part of an image asfigure, it cannot simultaneously alsobe part of the ground We can study the kinds of choicesthat the visual system makesPurdue UniversityChoices Generally, smaller areas seem to be part of figurePurdue UniversityChoices Symmetric things seem to be figuresProf. Greg Francis4PSY 310: Sensory and Perceptual ProcessesPurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scene We’ll try a demonstration of this effectPurdue UniversityFamiliarity What you decide is figure or ground also depends on how you tendto interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a sceneProf. Greg Francis5PSY 310: Sensory and Perceptual ProcessesPurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a sceneNow everybody for the next slideProf. Greg Francis6PSY 310: Sensory and Perceptual ProcessesPurdue UniversityFamiliarity What you decide is figure or ground also depends on how youtend to interpret a scenePurdue UniversityObjects One way to study object perception is to measure how long ittakes us to identify different objects Sounds easy, but it can be difficult One common technique is the visual search paradigm A target may or may not be presented among a set of distracters Your task is to identify whether the target is present or not The experimenter varies the properties of the target anddistracters To identify what properties/features of objects are conducive to thetaskPurdue UniversityCogLab feature search Few distracters -easyPurdue UniversityCogLab feature search Many distracters - still easyPurdue UniversityConjunctive search Few distracters - fairly easyPurdue UniversityConjunctive search Many distracters - difficultProf. Greg Francis7PSY 310: Sensory and Perceptual ProcessesPurdue UniversityVisual search Group CogLab data (93 observers)Purdue UniversityVisual search Interpretation Feature searches do not vary with number of distracters» feature searches “pop-out,” the distracters are irrelevant Feature present = feature absent» no searching is required, the reaction time just reflects the timeto make the response Conjunctive searches increase RT with number of distracters» must search individual items and compare with target to see if itmatches, each comparison takes time (attention) Conjunctive absent > conjunctive present» must search through more items (on average twice as many) forabsent than present searches There are thousands of variations of visual search experimentsPurdue UniversityInterpretation Feature maps: color, shape Feature search can identify target within either feature mapShapeColorbluegreencornersarcsPurdue UniversityInterpretation Feature maps: color, shape Feature search can identify target within either feature mapShapeColorbluegreencornersarcsPurdue UniversityInterpretation Feature maps: color, shape Conjunctive search cannot identify target within either feature map