Fichas
Objects & Scenes
Fichas del conjunto (45)
- 1What is the inverse projection problem? How is it completed?This is the task of taking an image from the retina and understanding what is truly in the environment. This is difficult because, for any image projected on the retina, there are an infinite number of objects which may cause it. This is because 2D images are ambiguous. People must use unconscious inference to determine the likely cause of the image. This also allows us to identify objects which are partially hidden or blurred.
- 2What is the Gelstalt Movement? What phenomenon led to the development of this?This is the idea of considering conscious experience globally. This contrasts with the idea of structuralism (circa late 19th Century), or the idea of understanding perception strictly via bottom-up 'elementary sensations'. Gestalt theory views perception as being neither direct nor local.
This is evidenced by colour constancy, or the maintenance of colour despite global tints, showing the importance of context.
An observation of apparent motion (the way people can see flickering dots as moving) led to the development of this approach. This cannot be explained by elementary sensations, because we do not sense the movement. This phenomenon is the effect of knowledge, expectations, attention, and motivation. - 3What is structuralism?This is an out-dated way of studying perception, which sought to explain perception by breaking it down into 'elementary sensations'. It focused solely on bottom-up processing.
During studies, subjects observed and described visual stimuli, and these descriptions were used to determine 'elementary sensations'.
However, there was disagreement between labs on what these fundamental properties of vision were (e.g. shapes, colours, etc.), and there were an unmanageable number of them (~40,000). - 4What are the Gestalt grouping principles?Simplicity/Pragnanz
Similarity
Common region
Continuity/good continuation
Proximity
Common fate
Connectedness
Synchrony
Meaning - 5What is the simplicity/pragnanz grouping principle?This describes the way people perceive objects as being occluding or attached, rather than being separate objects. This also includes organising objects in a systematic array as being related, rather than as a meaningless collection.
This can also includ symmetry. - 6What is the similarity grouping principle?This describes the way people perceive objects which are similar as being part of a discernible pattern.
- 7What is the common region grouping principle?This describes the way people perceive objects within the same discrete region as being grouped.
- 8What is the continuity/good continuation grouping principle?This describes the way people perceive objects as having smooth, continuing edges even when occluded.
- 9What is the proximity grouping principle?This describes the way people perceive objects which are closer together as grouped.
- 10What is the common fate grouping principle?This describes the way people perceive objects which are moving together as being grouped together.
- 11What is the connectedness grouping principle?This describes the way people perceive objects which are linked as being grouped.
- 12What is the synchrony grouping principle?This describes the way people perceive objects which are moving in a particular direction in contrast to the direction of other objects as being grouped together. This is a more complicated version of common fate.
- 13What is the meaning of grouping principle?This describes the way people perceive objects are perceived, according to expectations or prior knowledge of what is present.
- 14What is image segmentation? Why is this difficult, and what methods are used to achieve it?This is the process by which different areas or objects are separated by the perceptual system, allowing the identification of discrete objects with borders.
Neurons in V1 detect countless edges in different scenes, some being changes in light, or changes in object. The perceptual system has to determine which edge bits should be joined, where the bounding contours of each object are, and which side of a contour is object and background.
This is solved through contour grouping and boundary identification. - 15What is contour grouping?This is way the perceptual system determines where the bounding contours of an object are according to the proximity and flow of the edges.
- 16What is boundary identification?This describes the way the perceptual system uses changes in texture and light to determine where objects end.
- 17What are some nonclassical figure-ground cues?Familiarity with a specific shape as being an object makes us more likely to determine its silhouette to be figure. If inverted, however, this is disturbed, so it may be considered the ground.
If the texture and the border are moving together, it's seen as the figure. If the texture is moving alone, it's considered the ground.
If one section is blurred and the other is not, attention is paid to the border between these objects. If the border is not blurred, the clear section is seen as a figure. If the - 18What are examples of nonclassical figure-ground cues?A person's familiarity with a particular shape will lead them to see it as the figure. This is disturbed when the shape is inverted.
The movement of a texture will appear to be in the background unless the border moves with it, as that indicates that the moving object "owns" the border, and is thus the figure.
Familiarity with natural objects also leads us to perceive bottom-heavy objects as figures.
Blur can also indicate ground.
External edge cues can be very strong influencers over perception. Indicators of texture and shape (e.g. shading indicating roundedness) can overcome other cues of size, colour, and convexity. - 19What is viewpoint invariance?This is ability to recognise an object despite seeing it from a different viewpoint - you don't need to have a specific image projected on the retina in order to identify the object.
- 20What are accidental and non-accidental viewpoints?Non-accidental viewpoints are typical, familiar images which are recognised as a particular object. They show view-invariant features (e.g. the handle on a mug) which can be used to identify it.
Accidental viewpoints are unusual perspectives which often hide view-invariant features (e.g. seeing the face of a bug up close). - 21How is the fusiform gyrus involved our understanding of faces?This area contains the fusiform face area (FFA), which is used for facial recognition. Damage causes prospagnosia, or the inability to recognise faces.
- 22How is the subcortical region of the brain involved our understanding of faces?This area is used to identify fearful faces.
- 23How is the superior temporal sulcus (STS) involved our understanding of faces?This part of the brain is used to process the eye gaze and mouth movement of others.
- 24How is the frontal cortex involved our understanding of faces?This part of the brain is used to evaluate facial attractiveness.
- 25How is the amgydala involved our understanding of faces?This part of the brain is activated by the emotional aspects of faces.
- 26How is the processing of faces lateralised?The left hemisphere is activated by anything even vaguely face-like. The right only activates when observing faces.
- 27What is facial pareidolia?This is the human tendency to see faces in inanimate objects. This is an example of how top-down knowledge influences processing.
- 28Why are some objects more difficult to perceive than others? What can aid in perception?When there is little differentiating detail (e.g. objects which lack edges and grouping cues), bottom-up processing is hindered. However, this can be aided by top-down knowledge (e.g. knowing what should be in the scene).
- 29What is backward masking? And what has it indicated about perception?This is an experimental paradigm where the participant sees a scene very briefly, followed by a colourful but meaningless 'mask' stimulus. The participant is asked to categorise the scene.
These experiments have shown that this categorisation can happen very quickly, as expectations about particular objects quickly direct attention to important parts of the scene.
It also indicates that people perceive scenes globally, without paying attention to specific objects, during recognition. Even averaging all the categorised scenes still produces a categorisable image. - 30What is perceptual organisation?This is the process by which people use grouping to arrange the elements of the environment into coherent units or objects. This process works in conjunction with segmentation to understand a scene.
- 31What are illusory contours? How do they support the Gestalt approach?This describes the way non-existent contours are perceived when viewing an arrangement of shapes with certain sections removed. There are no sensations of the contours, and yet a shape is perceived, showing that the sum of perception is greater than the combined parts, or elements.
- 32What cues indicate the presence of a figure?- occluding the ground
- 'owning' the border (the border defines the edge of the figure, not the background)
- more memorable
- convex (coming back in itself) (these cues are stronger when there are multiple objects)
- lighter
- symmetrical
- smaller
-closed
- appear to be in front
- make up the lower section of the scene (reflecting the way natural objects are typically below the horizon) - 33Does recognition occur before, after, or while determining the figure? How do we know?Recognition either occurs before and while determining the figure.
In a study, participants were more likely to consider a familiar shape the figure than an unfamiliar shape, indicating the influence of top-down knowledge (recognition) over figure identification. - 34What is the difference between an object and a scene?A scene is a real-world environment containing multiple objects, which are organised in a meaningful way in relation to each other and the background.
In short, an object is acted on, while a scene is acted within. - 35In an experiment, participants viewed images for times varying from 27-500 ms, and were told to describe them. What were the findings?If they saw the scene for less than 67 ms, they could only perceive dark and light patches. At 67 ms, they could identify large objects, such as a table or person. At 500, they were able to recognise smaller objects and details.
This shows that, while looking at a scene, we first acquire the gist of the scene, followed by the details. - 36What are global image features?These are holistic, rapidly-perceived properties of the environment. They depend on the person's familiarity with regularities in the environment. Global image features include:
degree of naturalness (determining whether natural or man-made according to the curvature of contours)
degree of openness (beaches being very open, forests not. Indicated by few objects and visible horizon)
degree of roughness (smooth scenes have fewer and smaller objects)
degree of expansion (in the third dimension, indicated by the convergence of parallel lines. Depends on the viewpoint)
colour (e.g. characteristic colours of beaches, forests, etc.) - 37What is the significance of regularities in the environment?Experience seeing regularities in the environment allow a person to utilise global image features, which permit the holistic and rapid perception of scenes.
- 38What are examples of physical regularities in the environment?Both man-made and natural environments include more horizontal and vertical angles than oblique ones.
When one item partially occludes another, the occluded item continues on the other side.
Light-from-above assumption: the assumption that the light source is above the viewer, as well as cues from shading, help the person determine the shape of an object (including convexity/concavity). - 39What are semantic regularities? What research has been done regarding their influence?These are characteristics associated with different types of scenes, leading to the development of scene schemas, which describe what features a particular scene typically includes.
In a study, a particular type of scene was briefly flashed (e.g. a kitchen), followed by three objects. Two of these were inconsistent with the scene (e.g. a drum), and one may have fit in the scene (e.g. a toaster). Participants could identify the consistent objects with 80% accuracy, but only showed 40% accuracy on inconsistent objects.
In another study, participants saw blurred pictures, and needed to identify a particular blob. Throughout the different scenes, participants identified the blob differently to be consistent with their scene schema, but the blob itself was always the same. This shows the way that we're not always aware of the specific information we're using during perception. - 40Describe the principle of Bayesian inference.This principle suggests that we combine our previously-held assumptions (priors) with what additional information is supplied by the enviornment (the likelihood) to determine the probability of a particular perception (outcome) being correct.
For example, if we have the prior that books are rectangular, then we see a book, and the retinal image of the book supplies further evidence (likelihood) that the book is rectangular, then the probability of the outcome of the book being rectangular is high. Thus, our perception of the book being rectangular is strengthened. - 41What other name is the parahippocampal place area (PPA) called by? Why?It's also called the parahippocampal cortex (PHC), because there are disputes regarding its function.
Scenes cause more activation than buildings, but buildings cause more activation than general objects.
Space defining objects (such as a bed) cause more activation than space ambiguous objects (such as a fan).
Scenes which gave an impression of a 3D space, even if they contain few objects (such as a wide open field) also activates it. - 42Describe the way perception of faces develops in newborns over time.Newborns can only make out stark light and dark contrasts. 8-week-olds can make out a face-like image. 3-4-month-olds can distinguish between happy, surprised, angry, and neutral faces.
- 43The preferential looking technique has been used to determine whether an infant can recognise their mother. What were the results of this experiment? What change to the mother's appearance altered this result?Infants looked at their mother's face 63% of the time, suggesting they recognise her. However, by covering up the mother's hairline with a pink scarf, this effect goes away, showing the significance of this contrast.
- 44How does experience affect development of facial recognition?A study found that in three-day-olds, those who had more time exposed to their mother's face were more likely to prefer looking at her.
- 45Why does recognition of facial expressions take until adolescence or young adulthood to fully develop?Because the FFA develops slowly.
Información
- Descripción
- Perception Fichas sobre Objects & Scenes, creado por Jamie Otten el 27/02/2019.
- Creado por
- Jamie Otten
- Visitas
- 0
- Valoración