Psychology Notes - CH5-6 Regularities and Inference

**[CH5 contd.]** [Regularities in the environment] how do we form long-term/prior knowledge? Perception is influenced by 2 types of regularities in the environment: **1)Physical regularities:** Regularly occurring physical properties of the environment. Comes from physics of the environment. E.g...

**[CH5 contd.]** [Regularities in the environment] how do we form long-term/prior knowledge? Perception is influenced by 2 types of regularities in the environment: **1)Physical regularities:** Regularly occurring physical properties of the environment. Comes from physics of the environment. E.g. there are more vertical and horizontal orientations in the environment than oblique orientations. people can perceive horizontals and verticals more easily than other orientations. E.g. light comes from above assumption we usually assume that light is coming from above, because light in the environment (such as sun and most artificial light) usually comes from above. -Our perception off illuminated shapes is influenced by how they are shaded, combined with the brain's assumption that light is coming from above. -One of the reasons humans are able to perceive and recognize objects & scenes so much better than computer-guided robots our perceptual system is adapted to respond to physical characteristics of our environment, such as the orientation of objects and the direction of light. But this adaptation goes beyond physical characteristics. It also occurs because we have learned about what types of objects typically occur in specific types of scenes. **2)Semantic regularities:** Characteristics associated with activities that are common in different types of scenes. Semantic refers to the meaning of a scene, it is often related to what happens within a scene. E.g. food preparation, cooking, eating happens in a kitchen; buying tickets, checking luggage happens in an airport. -Our visualizations contain info based on our knowledge of different kinds of scenes. This knowledge of what a given scene typically contains is called a *scene schema.*  -First presented a context of a scene on the left and then briefly flashed one of the target pictures on the right. -Observers correctly identified A 80% of time, but correctly identified B and C only 40% of the time. they were using their knowledge about kitchens to help them perceive the briefly flashed loaf of bread. E.g. multiple personalities of blob the blob is perceived as different objects depending on its orientation and the context within which it is seen. -Context affects how we perceive things if something is ambiguous, we incorporate our prior knowledge/context in order to interpret the image. -Although people make us of regularities in the environment to help them perceive, they are often unaware of the specific info they are using. [The role of inference in perception] -People use their knowledge of physical and semantic regularities to infer what is present in a scene. -We have seen that retinal ambiguity means that a particular pattern of stimulation on the retina can be caused by many different possible objects in the environment. **Helmholtz's Theory of Unconscious Inference:** How does the perceptual system decide that the image on the retina (a) is created by (b)? Answer: **Likelihood principle:** We perceive the object that is *most likely* to have caused the pattern of stimuli we have received. This judgment of what is most likely occurs, according to Helmholtz, by a process called **unconscious inference:** our perceptions are the result of unconscious assumptions/inferences that we make about the environment based on our experiences. Thus, we infer that it is likely that (a) is a rectangle covering another rectangle because of experiences we had with similar situations in the past. -Perception is an unconscious inference process. We combine perceptual input/info with our prior knowledge to get inference/interpretation. [Bayesian Inference] Our estimate of the probability of an outcome is determined by 2 factors: **1)Prior:** Our initial estimate of the probability of an outcome. **2)Likelihood:** The extent to which the available evidence is consistent with the outcome. Example: a)Mary's priors in her head for 3 types of health problems. She believes that having a cold or heartburn is likely to occur, but having lung disease is unlikely. -Her friend Charles has a bad cough, she guesses that three possible causes could be cold, lung disease, or heartburn. b)She does some research and finds that coughing is often associated with having either a cold or lung disease, but isn't associated with heartburn. c)This additional info (likelihood) combined with Mary's prior to produce the conclusion that Charles probably has a cold. -People start with a prior, then use additional evidence to update the prior and reach a conclusion. [Bayesian inference in object perception] -Inverse projection problem occurs because of a huge no. of possible objects could be associated with a particular image on the retina. Problem is how to determine what is out there that is causing a particular retinal image. E.g. -Prior your initial belief that books are rectangular. -Likelihood provided by additional evidence such as the book's retinal image, combined with your perception of the book's distance and the angle at which you are viewing the book. -If this additional evidence is consistent with your prior that the book is rectangular, the likelihood is high and the perception "rectangular" is strengthened. -Conclusion shape is a rectangular. -While the retinal image is still the starting point for perceiving the shape of the book, adding the person's prior beliefs reduces the possible shapes that could be causing that image. -It occurs automatically and rapidly. -Because Bayesian inference provides a specific procedure for determining what might be out there, researchers have used it to develop computer vision systems that can apply knowledge about the environment to more accurately translate the pattern of stimulation on their sensors into conclusions about the environment. [Connecting neural activity and object/scene perception] -We have seen that *looking at* a face or place causes activity in the FFA or PPA. But is there any evidence that *perceiving* a face or place is associated with activity in these areas? -In normal everyday perception, our 2 eyes receive slightly different images because they are in 2 slightly different locations. These two images, however, are similar enough that the brain can combine them into a single perception. -But if the two eyes receive totally different images, brain cannot combine these two images and a condition called **binocular rivalry** occurs observer perceives either left-eye image or the right-eye image, but not both at the same time. [Experiment:] Researcher used binocular rivalry to connect perception and neural responding by presenting a picture of a person's face to one eye and a picture of a house to the other eye and having observers view the pictures through colored glasses. -The colored glasses caused the face to be presented on the left eye and the house to the right eyes. Because each eye received a different image, binocular rivalry occurred while the images remained the same on the retina, observers perceived just the face or just the house, and these perceptions alternated back and forth every few seconds. -The subjects pushed one button when they perceived the house and another button when they perceived the face, while fMRI activity was recorded in PPA and FFA. -When perceiving the houseactivity increased in PPA, decreased in FFA. -When perceiving the faceactivity increased in FFA, decreased in PPA. -Even though the images on the retina remained the same throughout the experiment, activity in the brain changed depending on what the person was experiencing. -Relationship btw perception and brain activity exist. [Neural Mind Reading] -Measuring the person's brain response and determining the stimuli that generated that response. -Using a neural response (usually brain activation measured by fMRI) to determine what a person is perceiving/thinking. -fMRI measures the activity in **voxels** small cube-shaped volumes in the brain about 3mm on a side. The pattern of voxels activated depends on the task and the nature of the stimulus being perceived. -Viewing a different orientation activates another pattern of voxels. a)viewing an oriented grating like the one on the left causes a pattern of activation of voxels. The cubes in the brain represent the response of eight voxels the differences in shading represent the pattern of activation of the orientation being viewed. b\) Results of Tong's experiment for 2 orientations. The gratings are the stimuli presented to the observer. The line is the orientation predicted by the decoder. The decoder was able to accurately predict the orientations of all eight of the gratings tested. -Basic procedure for neural mind reading: a\) The relationship btw orientation and the voxel pattern is determined by measuring the brain's response to a number of orientations. --As subject looks at different orientations, fMRI is used to determine voxel activation patterns for each orientation. b\) Then these data are used to create a "decoder" program that can determine orientation based on the voxel activation pattern collected in (a). c)Finally, the decoder is tested by measuring different brain activation as a person is looking at different orientations. This time using the decoder to predict the orientation the person is perceiving. --As a subject looks at an orientation, the decoder analyzes the voxel pattern recorded from the subject's visual cortex. Based on this voxel pattern, the decoder predicts the orientation that the subject is observing. If this works, it should be possible to predict what orientation a person is looking at based on their brain activation alone. **Decoding** You create a computational model that maps input to output. This model learns what kind of perceptual input creates what kind of activity in the brain. -You provide neural activity to the model and ask "what is this"? -Models are like a baby's brain. You show/teach them apples and oranges for example. To teach how to distinguish them, you show the baby examples of oranges and apples multiple times. Buraya defterdeki örneği çiz: -Scientists look at visual cortex (sensitive to orientations and edges). E.g. whenever you see a 45-degree orientation we see a [pattern] in the brain. AI model learns this brain activity pattern (similar to sensory coding). Voxels give certain patterns of responses. -If you just look at only 1 voxel activity at a time, you cannot get the orientation that is viewed. Instead of looking at single voxel activity, we look at pattern of activity in voxels. -You can decode the orientation with pattern of voxel activity. can pick up subtle differences even with fMRI. **[CH6: Attention]** **Attention:** The process of focusing on specific objects while ignoring others. -The act of attention enhances the processing of that object and therefore our perception of the object. -There are lots of stimuli in the real life, it is impossible for us to perceive everything at once -Our perceptual system has a limited capacity for processing info, so it needs to select certain things in the environment for enhanced processing. -Attention is a filter mechanism that filters out the stimuli that are relevant to us. -Attention is a top-down processaffects how we perceive things. [Visual scanning] -One of the mechanisms for selecting certain things in the visual environment for enhanced processing is **visual scanning** looking from one place to another. This scanning is necessary because there is only one place on the retina that creates good detail vision. **-Fixation:** Briefly pausing on an object. Provides us the opportunity to focus on a particular person/object so that we recognize them. **-Saccadic eye movement:** rapid jerky movement from one fixation to the next. These eye movements allow us to shift our attention and focus to other people and objects in a scene. Scanning involves **Overt attention:** attention that involves looking directly at the attended object. Can be studied by tracking eye movements. **Covert attention:** Attention without directly looking (without fixating an object). E.g. enables us to monitor the actions of a person sitting near us without staring. -We are constantly shifting our overt and covert attention to monitor what is going on in our environment. [What directs our attention?] Attention can be caused by: **1)Involuntary processes**, aka *exogenous attention*, aka *bottom-up attention*: in which stimuli that stand out capture our attention. We don't have the goal for looking at something but it (external stimulus) captures our attention. **2)Voluntary processes**, aka *endogenous attention*, aka *top-down attention*: in which attention is guided by our goals and intentions. **Visual salience** bottom-up attention -Some things in the world draw our attention because they stand out against their backgrounds. Scene regions that are markedly different from their surroundings, whether in color, contrast, movement, or orientation, are said to have **visual salience.** -Properties of a stimulus grab attention, seemingly against a person's will---**attentional capture.** Even though attentional capture can distract us from what we want to be doing, it is an important means of directing attention. Conspicuous stimuli like sudden movement or loud sound can capture our attention to warn us of something dangerous like an animal or object moving rapidly toward us. **-Saliency map**: reveals which regions are visually different from the rest of the scene. Regions of greater visual salience are denoted by brighter regions in the saliency map. [Cognitive factors] **Scene schemas** top-down attention Attention is influenced by **scene schemas:** an observer's knowledge about what is contained in typical scenes. Observers looked longer at the printer in (a) than the pot in (b) because a printer is less likely to be found in a kitchen. -The fact that people look longer at the things that seem out of place in a scene means that attention is being affected by their knowledge of what is usually found in the scene. -We have certain expectations of what a scene includes. If our perception mismatch with our scene schemes, we look longer. **Observer interests and goals:** -Attention can also be influenced by a person's goals. What you attend to changes depending on the goals. -Participants' eye movements were recorded while they viewed a painting. patterns of eye movements depended on the task participants were given. Eye movements were concentrated on faces when the task was to determine ages, on bodies when the task was to remember clothing, and were more evenly distributed across the painting when the task was to remember the positions of all the people and objects. -As people's intentions and tasks change, they will change how they focus their attention on a scene. **Task-related knowledge:** top-down attention -Our goal in viewing scenes isn't simply to perceive and understand objects and their arrangements, it prepares us for action. -People are shifting their attention from one place to another as they are doing things. -Timing of when people look at specific places is determined by the sequence of actions involved in the task. -Person's eye movements were determined primarily by the task. The person rarely fixated on objects or areas that were irrelevant to the task, and eye movements and fixations were closely linked to the action the person was about to take. -Eye movements occur just before we need the info they will provide. -Task-related knowledge gives us direction for where to attend. Task-related knowledgemore dynamic Scene schememore static -In both cases, you have a prior knowledge about a scene or activity. We use these previous experiences to direct our attention. -Difference between 2: scene schemes are more static and task-related knowledge is an activity that takes some time and series of events. [What are the benefits of attention?] -Attention enhances our: Response to objects (we respond faster to things that are located where we attend) Perception of objects (make it easier to see an object) Physiological responding (neural response to objects) [Attention speeds responding] -Paying attention informs us about what is happening at a location, and also enables us to respond more rapidly to anything that happens in that location. **Locations** **Spatial attention:** Attention to a specific location. [-Posner paradigm:] Subjects kept their eyes stationary throughout the experiment, always looking at the + in the display. They first saw an arrow cue indicating on which side of the target a stimulus was likely to appear. The subjects' task was to press a key rapidly when a target square was presented off to the side. -Valid trialsquare appears on the side indicated by the arrow que. -Invalid trialarrow cue indicated that target was going to be presented on one side but it actually appeared on the other side. -Results: subjects reacted more rapidly on valid trials than on invalid trials. info processing is more effective at the place where attention is directed. -Attention improves processing when directed toward a particular location. -The problem is you are making the subject not only attend but also expect for object to appear there confound of expectation. **Objects** -Attention can enhance our response to objects and when attention is directed to one part of an object, the enhancing effect of that attention spreads to other parts of the object. -As subjects kept their eyes on the +, one end of the rectangle was briefly highlighted this was the cue signal that indicated where a target, a dark square, would probably appear. -The subjects' task was to press a button when the target was presented anywhere on the display. -Subjects responded most rapidly when the target was presented at A, where the cue had been presented. -subjects responded more rapidly when the target was presented at B than when it was presented at C. B's advantage occurs because it is located *within the object* that was receiving the subject's attention. (grouping principle in gestalt) -Attending at A causes the maximum effect at A, but the effect of this attention spread throughout the object so some enhancement occurred at B as well. -When a person fixates on a location on an object, the entire object is selected for attentional processing. [Attention affects perception] -Carrasco et al. were interested in determining whether attention affected the perceived contrast btw the bars, where perceived contrast refers to how different the light and dark bars appear. Their hypothesis: attention would cause an increase in the perceived contrast of the gratings. -Subjects were instructed to keep their eyes fixed on the small dot at all times. -A cue stimulus was flashed for 67 ms either on the left or on the right. Subjects were told that this cue had no relation to the stimuli that followed it. -A pair of gratings was flashed for 40 ms. The gratings were tilted in different locations (one to the left, one to the right). -The contrast btw the bars of the gratings were randomly varied from trial to trial, sometimes contrast of the right was higher, sometimes left was higher, and sometimes two grating were identical. -The subjects' task was to indicate on each trial whether the grating stimulus with the greatest contrast was tilted to the left or to the right. -When 2 gratings were different attention-capturing dot had no effect. -When 2 grating were physically identical subjects were more likely to report the orientation of the one that was on the same side as the flashed cue. The one that received attention appeared to have more contrast. -A variety of other perceptual characteristics are affected by attention. E.g. attended objects are perceived to be bigger, faster, and more richly colored. [Attention can influence physiological responding] **Attention to objects increases activity in specific areas of the brain** -Subjects saw a face and a house superimposed, presented to both eyes so there was no binocular rivalry. Instead of letting rivalry select the image that is visible, experimenters asked subjects to direct their attention to one stimulus or the other. -For each pair, one of the stimuli was stationary and the other one was moving slightly back and forth. -When looking at a pair, subjects were told to attend to either the moving or stationary house, the moving or stationary face, or the direction of movement. As they were doing this, activity in their FFA and PPA was measured. -Results: Attending to moving/stationary face caused enhanced activity in the FFA and attending to the moving/stationary house caused enhanced activity in the PPA. -Attending to different types of objects influences the activity in areas of the brain that processes info about the type of object. **Attention causes changes in the relationship between activity in different areas of the brain** -Researchers demonstrated this by recording a response called the local field potential (LFP) from the monkey's cortex. LFPs reflect electrical signals from thousands of neurons near the electrode. -Researcher presented a monkey that was looking at a fixation dot with two stimuli and recorded from 3 locations in the monkey's brain. Locations A and B in area V1, in the occipital cortex, and location C in area V4, in the temporal cortex -Both A and B send signals to C. -It is found that Stimulus 1 caused an LFP response at location A in the cortex, and also at location C (because A sends signals to C). Stimulus 2 caused responses at locations B and C. -What happens when the monkey, while keeping its eyes on the fixation dot, directs its attention to one of the stimuli? -Paying attention to stimulus 1 caused little change in the size of the response recorded at A, and shifting attention to stimulus 2 had little effect on the responses recorded at B. -The effect of attention isn't on the *size* of the responses recorded from V1, but on the *relationship* between the responses recorded from V1 and V4. b\) There is little relationship btw the two responses. c\) Responses at A and C have become synchronized with each other. -They used a statistical procedure to determine a measure called **coherence:** indicated the degree to which two signals are synchronized. Coherence measure that tells you how the activity btw regions is coherent/similar. High coherence regions communicate with each other. -Results indicate that when the monkey was paying attention to stimulus 1, there was a synchronization btw the LFPs recorded from A and C. -When monkey shifted its attention to stimulus 2, the synchronization btw A and C vanished and was replaced by synchronization btw B and C. -Attention enhances the communication btw neurons in areas V1 and V4, by causing these neurons to fire with similar patterns. [Attention and experiencing a coherent world] **Binding:** process by which visual features (e.g. color, form, motion, location) are combined to create our perception of a coherent object. -When a person observes a red ball roll by, cells sensitive to the ball's shape fire in his IT cortex, cells sensitive to movement fire in his medial temporal cortex, and cells sensitive to color fire in other areas. But person doesn't perceive the ball as a separated shape, movement, and color perceptions. He experiences an integrated perception of a rolling red ball with all of its features bound together in a coherent whole. **Binding problem:** How an object's individual features become bound together? **Feature integration theory:** -The first step in object processing is the **preattentive stage:** occurs before we focus attention on an object. This stage is automatic, unconscious, and effortless because attention isn't involved. In this stage, the features of objects are analyzed independently in separate areas of the brain and are not yet associated with a specific object. -These independent features (e.g. color, form, motion) are then combined in **focused attention stage**. When the man focuses his attention on the object in front of him, the independent features derived in the preattentive stage become linked together so that he becomes consciously aware of a red ball rolling to the right. **Divided attention tasks** subjects are given multiple tasks to complete at once. The task was to report the numbers and report the shapes at each location. Subjects had to divide their attention across two tasks: identifying number and identifying shapes. their ability to focus attention on the shapes is reduced. What did subjects report seeing? Objects that were made up of a combination of features from 2 different stimuli. E.g. red circle or green triangle. **Illusory conjunctions:** Combinations of features from different stimuli. This can occur even if the stimuli differ greatly in shape and size. -According to FIT, these illusory conjunctions occur because the divided attention task reduces subjects' availability to focus their attention on the shapes, and this allows independent visual features to be combined incorrectly. -When subjects were instructed to ignore the black numbers and focus all of their attention on the four shapes, illusory conjunctions were eliminated and all of the shapes were paired with their correct colors. [What happens when we don't attend?] -Not attending can cause us to miss things even if we are looking directly at them. **Inattentional blindness:** Subjects can be unaware of clearly visible stimuli if they aren't directing their attention to them. Experiment: participants were shown a video where 2 teams are playing basketball, one team was dressed in white. Subjects were told to count the number of passes that white team made. They focused their attention on the team wearing white. After a while a person in gorilla suit walked through the game for 5 seconds in the video. -46% of the observers failed to report that they saw the gorilla. -When people are attending to one sequence of events, they can fail to notice another event, even it is right in front of them. **Change blindness:** Subjects were presented one picture, followed by a blank field, followed by the same picture but with an item missing, followed by a blank field, followed by the original picture and so on. -The pictures were alternated in this way until observers were able to determine what was different about them. This difficulty in detecting changes in scenes is called change blindness. [Is attention necessary for perceiving scenes?] **Evidence that perception can occur without attention** -People can identify the type of scene in a picture (e.g. forest, seashore) after seeing the picture for less than a ¼ of a second. -Dual-task procedure: subjects are required to carry out simultaneously a central task that demands attention and a peripheral task that involves making a decision about the contents of a scene. -Subjects looked at the + on a fixation screen and then saw the central stimulus, an array of 5 letters. On some trials, all of the letters were the same; on other trials, one of the letters was different from the other 4. The letters were followed immediately by the peripheral stimulus---either a disc that was half red and half green or a picture of a scene---flashed for 27 ms at a random position on the edge of the screen. -Subjects' central task: indicate if all of the letters in the central stimulus were the same. -Peripheral task: indicate whether the scene contained an animal or whether the colored discs were red-green or green-red. -Even though subjects had to keep their attention focused on the letters in the middle in order to carry out the central letter task, their performance was 90% on the peripheral picture task, but only 50% on the peripheral colored-disc task. -Properties of scenes can be perceived with little or no attention. **Evidence that perception requires attention** -Researchers wondered if the central letter task (previously mentioned task) did not distract attention enough. To test this idea, they created a letter-number task in which a series of letters and numbers were rapidly flashed (E.g. GNW4AY5T) and subjects indicated how many numbers they saw. -While they were doing this central task, subjects indicated if a picture flashed rapidly off to the side contained an animal or a vehicle. -When the peripheral animal-vehicle task was presented alone, without the central task, subjects were correct 89% percent of the time. However, their performance dropped to 63% when their attention was distracted by the central number-letter task. perception of natural scenes does require attention. -Perhaps, there are some aspects of scene perception that require attention and some that don't. [Distraction] **Task-irrelevant stimuli:** Stimuli that don't provide info relevant to the task at hand. Can potentially decrease our performance of a task. e.g. computer pop-ups. -Amount of distraction depends on properties of the distracting stimulus. However, the effect of a potentially distracting stimulus also depends on the characteristics of the task. -If the task is easytask-irrelevant stimuli have an effect on performance -If the task is hardtask-irrelevant stimuli have little or no effect on performance. [Experiment:] The subjects' task was to respond quickly when they identified a target, either X or N, in displays. Subject pressed one key if they saw the X and another key if they saw the N. -This task is easy for display like the one on the left (a) in which the target is surrounded by just one type of letter, like the O's. -Task becomes harder when the target is surrounded by different letters. -This difference is reflected in the reaction times, with the hard task resulting in slower reaction times than the easy task. -When a task-irrelevant stimulus is flashed off to the side, like the cartoon character, responding slows for the easy task but is affected only slightly for the hard task. [Attention and perceptual load] **Perceptual capacity:** refers to the idea that a person has certain capacity that can be used for carrying out perceptual tasks. **Perceptual load:** Amount of a person's perceptual capacity needed to carry out a particular perceptual task. -Low-load tasks easy, well-practiced tasks that use up only a small amount of the person's perceptual capacity. -High-load tasks difficult, not as well practiced tasks that use person's perceptual capacity more. -Amount of perceptual capacity that remains as a person is carrying out a task determines how susceptible the person is to being distracted by task-irrelevant stimuli. Circle represents a person's total perceptual capacity; shading represent the portion that is used up by a task. a\) only part of the person's resources are being used by a low-load task, leaving resources available for processing other stimuli that may be present. That is why we get distracted while doing easy task. b\) all of a person's perceptual capacity is being used by a high-load task, no resources remain to process other/irrelevant stimuli, so irrelevant stimuli can't be processed and they have little effect on performance of the task.