click below
click below
Normal Size Small Size show me how
percep 2B lec 4-6
| Question | Answer |
|---|---|
| Attention | taking in posession by mind one out of several simultaneously poss objects or trains of thought, withdrawal from some things in order to focus on others |
| Divided Attention | - Paying attention to more than one thing at a time - Eg driving |
| Selective Attention | Focusing on specific items + ignoring others |
| Exogeneous Attention | - Guided by environ - Something grabs attention (movement, sound, change in environ) |
| Endogenous Attention | - Guided by attender (internally) - What choose to pay atten to |
| Perception + attention | - Don’t perceive things without paying attention? - Exception = faces |
| Inattentional Blindness | - Don’t perceive object/event if don’t attend to it - Espesh for dynamic events - Gorilla ball throwing vid |
| Change Blindness | - When scenes chanfe btwn two discrete views = not sens to large diffs btwn scenes - Doesn’t happen if know location of change - Motion IS sens to changes (motion helps to pick it up) - Spot the difference |
| Continuity Error | - Movies + tv - Filmmakers make mistakes due to inattentional blindness + change blindness |
| Spotlight of Attention | - Analogy - Attention = spotlight - Things going on behind or around spotlight = unnoticed - BUT can with effort pay atten to things outside spotlight |
| Modularity | - Early vision = modular - Analysing sep feats separately ( Colour, orientation, motion) - Attention = glue that sticks them togeth? Cant have percep w/o atten |
| the binding problem | How brain integrates infro from diff sens feats into unified coherent percep of object/event |
| Feature Integration Theory (FIT) | 1. Pre attentive stage (features free floating) 2. Focused attention stage (features bound together) |
| Pre-attentive Stage | - Sense basic feats of scene (edge orientations,shapes, colours) - Feats exist in unbound state - O in rows of V = parallale search, the O pops out ( Orientation differences + differences in curvature) |
| focused attention stage | - Atten directed to spec locations in scene - Neurons to object only in their own receptive field = no ambiguitty, accurate representation of scene exists |
| parallel search | - O in rows of V - can be done quickly -"pop out" |
| Conjunction Search | - Looking for a conjunction of two features, shape + colour - find the red O (sea of red + purp Os and Ws) |
| Serial Search | Look at each object in a crowd, wordsearch scannign each letter in a row |
| Illusory Conjunction | - Errors in binding - Perceiving wrong comb of feats in a scene |
| Gamma Band Oscillation | Oscillations in brain synchronise when binding things together |
| Eye Movement | - involved in how atten directed - eye gaze alloc to right place right time = guide actions, simple salience model should be rejected - diffs for some ppl, autistic = less eye fixation |
| Monotropism | - theory of aut coming from comm - interest model - aut persons interests drawn in more strongly - fewer interests roused, attract more processing resources - harder tp deal w things outside interest (look at face + listen to what saying simult) |
| attentional enhancement | - atten enhances percep (respond faster to attended obj/loc, see things clearer/contrast enhan) - atten enhan physiolog resp (parietal cort firing rates of bis neurons incr if obj in recep field attended, neurons resp better in atten parts of vis field) |
| motion perception | - Delay + compare - Measure img at one place + time and then later at other place + time - Velocity = distance/time |
| Motion (Reichardt) Detector | - Detectors in vis syst - Intensity edge going from left to right, detected first by left and then from right detector - Detection from both detectors = comparator - Delay from left = causes detection to happen simultaneously |
| Apparent Motion | - Frame 2 = object in diff place from frame 1 - Object hasn’t actually moved, but see it moving from place 1 to place 2 - Eg = bell lights look like swinging but two diff lights lit up at diff points |
| Static Motion Illusion | - Kitaoka - Objects in frame not moving, move eyes around = looks like moving |
| Correspondence Problem | - How know what parts of image correspond to each other across diff scene frames |
| Nearest Neighbour Matching | - Assume features which match frame to frame are the ones closest by - Solution to correspondence prob - Matching pixels from one frame to nearest pixels in next frame |
| Common Fate | - Gestalt grouping rule - If things moving in same direc = tend to group them together -Eg = flocks of geese |
| Aperture Problem | - Area seen by 1 motion detector = small, causes problems determining direction of motion - barberpole |
| Barberpole Illusion | - Stripe pattern moving along pole appears to be moving in inconsistant direc with actual motion - Imposing vertical or horizontal appertures makes stripes look like moving diff direc |
| Optic Flow | - Pattern of retinal motion we see if moving toward or away from object - Gives info abt speed + direction of heading - Used in VR videogames, 5D theatre to give illusion of moving |
| Corollary Discharge | - differentiate btwn mvnt caused by external stim + self gend mosh - comparator comps ret mosh to efferent sig from eye (diff sigs = mosh perceived, mosh w eyes fixed or tracked mosh) (matched sigs = no mosh perceived, ret mosh due to mosh of eyes) |
| Area V5/MT-> monkey brain | - MT/v5 neurons are selective for direction of motion - Receptive fields built up by combining motion detectors together = get bigger cell w bigger receptive field responding to certain speed + direction of motion |
| Areas MST, FST -> monkey brain | - MSTd = dorsal, self motion - MSTv = trajectories, standard obj mmvnt - FST = actions |
| hMT+, V3 | - Human motion area - Parts have input from vestibular syst (balance, acceleration), gives info on self motion - V3 + Posterior Superior Temporal Sulcus = imp for combinations of form + motion |
| Biological Motion | - Attach lights to joints, film human moving in dark room = point light display - Simply a pattern of motion, recognised as moving human - can assume identity, emotion, gender - posterior superior temporal sulcus = sens to bio mosh |
| Biological motion in autism | - mckayet al 2012 - autistic + non - see point light display moving left, right or scrambled - understand funct connect - non autistic = access to diff pathways, areas which deal w mosh directly - aut = seeing as series of snapshots must connect |
| Wavelength | - electromag rad just below 400 + abv 700 nanometres = all colours - parti wavelength = partic clr sens - clr not carried in light itself, light has prop (wvlngth) = stirs up disposish to see clrs - clrs in objs = way reflect diff types of light |
| Benham’s Disc | - Black + white, when spin = see colours - As rotate, get oscillation freq which evokes colours - Diff rates for diff ppl |
| Spectral Reflectance | - Spectral reflectance graph - Proportion of light a surface/object reflects at each wavelength |
| Colour Diagnosticity | - Colour can be diagnostic for an object class - Yellow + curved = know it’s a banana - Can be uncanny if colour isnt right |
| Memory Colours | - hansen et al 2006 - grey bananas = bluer than grey courgettes - polystyrene fruit, viewing box light onto it - partips adjst light = fruit looks grey - grey 4 courg = more reddish (offset green) - grey 4 ban = blue (offst yellow) |
| Trichromacy | - Young-helmholtz - clr = outputs of 3 phys mechs w diff spectral sensitivities - clr match tests = any clrd light = matched by adjusting relative amnts of 3 prim lights (blue, red, green) - 3 cone types = 3 cone pigments |
| spectral sensitivity | - 3 cone types = 3 diff spectral sensitivites - short (s) = blue - medium (m) = green - long (l) = red |
| Metamer | Spectral distributions that is diff but looks the same colour |
| Opponent-Process Theory | - alt to young helmholtz by hering - nat pairing of clrs = black/white, red/green, blu/yell - clrs gend by opp processing (scales) - evidence = clr afterimages = see opp clr when adapt to partic clr |
| Opponent cells in LGN + ret | - combine trichrom by combining cone outputs - L-M = red/green (l oppos m) - (L+M)-S ) yel/blu (s oppos comb of l + m) |
| Opponent processing = ret ganglion cells | r/g or b/y centre surround |
| opponent processing: encoding along optic nerve | - achrom channel = L+M = M layers = parasol/M ganglion cells - red/green = L vs M = P layers = midget/P ganglion cells - blue/yellow = (M+L) vs S = K layers = bistratified/K gang cell |
| Forbidden Colours | Reddish green, yellowish blue |
| Double Opponent Neurone | - Partic type of cell, responds only to clr contrast - Exist in v1 - Basis of simult clr contrast - Neurons built = red-green opponent cells together = new cell responding in partic way in diff parts of receptive field |
| Simultaneous Colour Contrast | - Grey patch surrounded by red, see green in middle/grey - Surround of double opponent cell stimulated by background, centre expects there to be red |
| Unique Hue | - Wavelengths not perceived as combos, but as purely blue green or yellow - Represented in v1 |
| Object Colour | - Determined by spectral distribution of reflected light - Spectral power of illuminant - Spectral reflectance of object |
| Colour Constancy | - Colours stay constant under changes of illuminant, don’t change moving from daylight ot artifish light or at diff times of day - achieved thru chromatic adaptation, effects of context (clrs change based on clrs next to it + illuminant) + mem effects |
| Chromatic Adaptation | - Wavelength strikes retina for extended time, cones most sens to those wavelengths become temp less sens to those wavelengths - Wearing coloured glasses, adapt to seeing world w diff colour filter |
| Lightness Constancy | - Calibrate using lightest + darkest things we see - What seen as white + black is relative - Helps to distinguish shadows from material changes |
| “Colour as Material” Assumption | - Colour = best + most reliable indicator of change in material rather than illumination change - Lightness boundary = shadow - Colour boundary = material changed |